COLUMBIA LIBRAHIf S Uhhgii c HEALTH SCIENCES STANpARD HX641 42027 QP701 .Sw2 Nutrition investigat RECAP Columbia SSnitJers^ftp CoUege of ^fjpsiicians; anir burgeons; ILibvaxp TRANSACTIONS OF THE CONNECTICDT ACADEMY OF ARTS AND SCIENCES Incorporated A.D. 1799 VOLUME 1 6, PAGES 247-382 APRIL 1911 Nutrition Investigations on the Carbohydrates of Lichens, Algae, and Related Substances BY MARY DAVIES SWARTZ FROM THE LABORATORY OF PHYSIOLOGICAL CHEMISTRY SHEFFIELD SCIENTIFIC SCHOOL YALE UNIVERSITY NEW HAVEN, CONNECTICUT, U. S. A. YALE UNIVERSITY PRESS NEW HAVEN, CONN. 1911 5 w 3^ COMPOSED AND PRINTED AT THE WAVERLY PRESS By The Williams & Wilkins Company Baltimobe, U. S. a. CONTENTS. I. INTRODUCTION. PAGE Lichens, Algae, Tree Bark and Certain Tubers and Foodstuffs 253 II. HISTORICAL PART. Introduction 259 Cellulose 262 (a) Occurrence and Nature 262 (b) Occurrence of Cytases (Cellulases) 264 (1) In the Vegetable Kingdom 264 (2) In Lower Animals 266 (3) In Higher Animals .267 (c) Digestion and Utilization 268 (1) By Animals 268 (2) By Man 269 The Pentosans 272 (a) Occurrence and Nature 272 (b) R61e in Plant Physiology 274 (c) Occurrence of Pentosanases 275 (1) In the Vegetable Kingdom 275 (2) In Lower Animals 276 (3) In Higher Animals 276 (d) Digestion and Utilization 278 (1) By Animals 278 (2) By Man 278 The Galactans 282 (a) Occurrence and Nature 282 (b) Occurrence of Galactanases 284 (1) In the Vegetable Kingdom 284 (2) In the Animal Kingdom 285 (c) Digestion and Utilization by Animals and Man 285 The Mannans 289 (a) Occurrence and Nature 289 (b) Occurrence of Mannanases 291 (1) In the Vegetable Kingdom 291 (2) In the Animal Kingdom 292 (c) Digestion and Utilization by Animals and Man 293 The Levulans 295 (a) Occurrence and Nature 295 (b) Occurrence of Levulanases 296 (1) In the Vegetable Kingdom 296 (2) In the Animal Kingdom 297 (c) Digestion and Utilization by Animals 297 249 250 Contents The Dextrans 300 (a) Occurrence and Nature 300 (b) Occurrence of Dextranases 301 (1) In the Vegetable Kingdom 301 (2) In the Animal Kingdom 302 (c) Digestion and Utilization by Animals and Man 302 III. EXPERIMENTAL PART. Introduction 306 Chemical Investigations. General Methods 307 Pentosan Preparations 309 (a) Dulse {Rhodymenia palmata) 309 (b) Hawaiian Seaweeds 313 (1) Limu Lipoa (Haliseris 'pardalis) 313 (2) Limu Eleele {Enteromorpha intestinalis) 313 (3) Limu Pahapaha (Ulza lactxica, etc) 314 Galactan Pkeparations 314 (a) Irish Moss (Chrondus crispus) 314 (b) Hawaiian Seaweeds 316 (1) Limu Manauea {Gracilaria coronopifolia) 316 (2) Limu Huna {Hypnea nidifica) 316 (3) Limu Akiaki (Ahnfeldtia concinna) 316 (4) Limu Uaualoli {Gymnogongrus vermicularis Americana, etc) 316 (5) Limu Kohu (Asparagopsis sanfordiana) 316 (c) Slippery Elm (Ulmus fulva) 317 A Mannan Preparation — Salep {Orchis) 318 A Levulan Preparation — Sinistrin (from Scilla Maritima) 321 Summary 322 BACTERIOLOGICAL INVESTIGATIONS. Introduction 323 Trials with pure cultures of aerobes 324 Trials with mixtures of aerobes 325 Trials with anaerobes 327 Discussion and summary 328 PHYSIOLOGICAL INVESTIGATIONS. Introduction 331 Experiments with Enzymes 332 Parental Injections ' 332 (a) Methods and Technique 332 Contents 251 (b) Subcutaneous and Intraperitoneal Injections 335 (1) Dulse 335 (2) Irish Moss 336 (3) Salep 338 (4) Sinistrin 340 Feeding Experiments 342 (a) Methods and Technique 342 (b) Digestibility of Pentosans 344 (1) Dulse 345 (2) Limu Eleele 346 (3) Litnu Pahapaha 347 (4) Limu Lipoa 347 (c) Digestability of Galactans 348 (1) Irish Moss 349 (2) Limu Manauea 350 (3) Limu Huna 351 (4) Limu Akiaki 351 (d) Digestibility of Mannan 353 (1) Salep 354 Discussion and Summary 356 IV. CONCLUSIONS. V. BIBLIOGRAPHY. Lichens and Algae — Composition and uses 365 Cellulose 366 Pentosans 369 Galactans 373 Mannans 376 Levulans 379 Dextrans 381 This paper has been prepared from the author's dissertation submitted for the degree of doctor of philosophy, Yale University, 1909. Digitized by tine Internet Arciiive in 2010 witii funding from Open Knowledge Commons http://www.archive.org/details/nutritioninvestiOOrose I. INTRODUCTION. Lichens, Algae, Tree Bark and Certain Tubers as Foodstuefs. From the earliest times, the food of man has included lichens and algae, and even the tender branches and inner bark of certain trees and shrubs, such as elm, birch, pine, and the staff-tree or bitter-sweet {Celastrus scandens). When the bark of trees is so used, it is freed from cork and the hard outer rind; is cleaned, dried, mixed with more or less meal, and made into "bark bread," Such substitutes for bread are commonly resorted to only in northern lands where there is scarcity of cereal crops, or in other regions during periods of famine. Johnson (7) records that elm bark is so employed in some continental countries, and Dillingham (4) relates that certain tribes of North American Indians, 'in times of extreme dearth, were accustomed to keep body and soul together by boiling and eating the bark of the staff- tree.' Poulsson (17) states that in Finland and northern Russia, sphagnum mosses are similarly employed; and Schneider (21) agrees with these other writers, saying that in general lichens are used as articles of diet only in cases of special need, principally because all lichens contain a bitter principle, which not only gives an unpleasant flavor and is difficult to remove, but also exerts an irritating effect upon the digestive tract, causing inflammation. Nevertheless, in the northern parts of the Scandinavian Peninsula, where cereal crops are always scanty or uncertain, great interest attaches to two species of lichen widely distributed through Europe, and through Arctic and Antarctic regions: namely, Ceiraria islandica and Cetraria nivalis, which, as Poulsson (17) observes, 'have been considered nutritive and easily digestible since olden times. ' Cetraria islandica, whitened and freed from its bitter principle by washing with dilute alkali, is a rather appetizing substance; it has sometimes been used as a foodstuff by Polar navigators, and Dr. Hansteen, chief lecturer in the Agriciil- tural school at Aas, Norway, has gone so far as to prophesy that moss is destined to become the great popular food for the masses, because of its cheapness and nutritive properties. Of marine algae, many tons are gathered and eaten annually in various parts of the world, the largest quantities being consumed 253 254 Mary Davies Swartz, by the Japanese, Chinese, and Hawaiians. These algae are found in great variety and widely distributed. In Japan, the general name applied to them is "Nori," which is also given to several prepared products. According to H. M. Smith (23), the most important Japan- ese seaweed preparations are: "Kanten," or seaweed isinglass, made from various species of Gelidium, the principal one being Gelidium corneum, often adulterated with similar seaweeds; "Kombu" made from Kelps, especially numerous species of Laminaria, Arthothamnus, and Alaria; "Amanori," from species of Porphyra; and "Wakame," from Undaria pmnatifida. Kanten is used largely for food, in the form of jellies, and as an adju- vant of soups and sauces. According to H. M. Smith (23), it is also employed in foreign countries 'in jellies, candies, pastries, and many desserts, in all of which it is superior to animal isinglass.' It has recently also attained popularity as a therapeutic agent in chronic constipation, being sold under various trade names, either plain or impregnated with laxative drugs, as cascara or phenolphthalein.i Kombu enters into the dietary of every Japanese family, being cooked with meat, soups, etc., and also served as a vegetable, or made into a relish with Soy-bean sauce. Amanori is eaten fresh or else is chopped and sun-dried in thin sheets, which are toastsd over a fire before eating. The crisp amanori is crushed between the hands and dropped into sauces or soups to impart flavor; or broken into pieces, dipped in sauce and eaten alone. Sheets of amanori, spread with boiled rice and covered with strips of meat or fish, are rolled and cut into trans- verse slices, and take the place of the American sandwich. Wakame is eaten as a salad, or cooked like amanori. In Hawaii, edible algae are called "limu. " Of these there are over seventy distinct species used for food, more than forty being in general use (18). Tons of limu are gathered for eating in Hawaii annually, and large quantities are also imported from the Orient and San Fran- cisco. Some idea of the extent of their use may be gained from the following statement by Miss Reed (18): "Ancient Hawaiians prob- ably seldom ate a meal without some kind of limu, and even today no Hawaiian feast is considered quite complete without several varieties served as a relish with meats or poi."^ Since, with the exception of a few experiments reported by Oshima (15) and Saiki (20), there are no iCf. Galactans, p. 283. ^Poi is a thick paste made from the root of the taro plant, and takes the place of rice or bread in the native diet. Nutrition Investigations. 255 data upon the digestibility of marine algae, an investigation of some of these Hawaiian limu seemed highly desirable; and through the kindness of Miss Reed, a number have been obtained for this purpose. Their occurrence and uses will therefore be described in some detail.^ These limu are washed carefiilly after gathering, salted, and usu- ally broken, pounded, or chopped into small pieces. They may then be eaten uncooked, as a relish with poi, meats or fish; boiled with meats; put into soups for thickening or flavoring; or roasted with pig in a pit. Served raw and crisp, they take much the same place in the diet as our salads. Among the most popular varieties are Limu Eleele {Enter o- morpha of various species), Limu Kohu {Asparagopsis sanfordiana) and Limu Lipoa {Ealiseris pardalis) . Next in favor come Limu Ma- nauea {Gracilaria coronopifolia) , Limu Buna {Hypnea nidifica) and Limu Akiaki {Ahnfeldtia concinna). Limu Pahapaha (Ulva fasciata and Ulva lactuca) is widely distributed but not very popular. Limu Uaualoli {Gymnogongrus vermicularis americana and Gymnogongrus disciplinalis) is limited to certain islands, and hence not in such gen- eral use and favor as some of the others. Limu eleele is a great favorite, forming a part of every native feast. It is generally eaten uncooked, sometimes being dropped into hot gravy, broth or meat stews just before serving. Limu kohu is always poimded in cleaning to free it from bits of coral and soaked 24 hours in fresh water to remove the bitter iodine flavor. It becomes sHghtly fermented and acqmres a somewhat sour taste. Limu lipoa is popular on account of its penetrating spicy flavor, and is frequently used as a condiment, taking the place of sage and pepper in Hawaiian foods. Limu huna is especially prized for boiling with squid or octo- pus, though Umu manauea and Umu akiaki are often used as substi- tutes. These limus, as well as limu kohu, yield large amounts of mucilaginous extract on boiling, limu manauea being considered es- pecially fine for thickening chicken broth. Many of the seaweeds used in Hawaii and Japan occur also along the coasts of the United States and Europe, and are to some extent used as food in both regions. The very species of Gelidium from which the Japanese prepare their Kanten grow in abundance on our Pacific coast. Irish moss {Chondrus crispus), the "Tsunomata" of Japan, has long had considerable commercial value as a foodstuff in Ireland. In this country it is found from North Carolina to Maine, being especially abrmdant north of Cape Cod. After cleansing, cur- ^For fuller description see Reed (18). 256 Mary Davies Swartz, ing, and bleaching it is to some extent used for making blanc mange or a demulcent for coughs. Through the kindness of Dr. C. F. Lang- worthy, Nutrition Expert, United States Department of Agriculture, I have obtained the foUowing interesting data concerning the use of Irish moss, from the Journal of the South-Eastern Agricultural Col- lege, Wye, Kent (1): "Professor D. Houston, of the Royal College of Science, Dublin, has favored us with the following notes on this sub- ject : Chondrus crispus (carrageen, or Irish moss) is a seaweed plentifully distributed along our northern, western and southern coasts. It is gathered and sold to local chemists, who retail it, in some parts at all events, at 6d. per pound. It is used by- many people as an article of food in the west, and generally for colds, for which pur- pose it is boiled in milk. Several of my students tell me that it is used for feeding weak calves and with striking results, bringing about an alteration of condition within four days. One student tells me that in one case at his own farm a batch of twelve calves took a kind of wasting disease, and nine died; the other three on the verge of death were given this plant, and all three recovered. It is prepared by putting one pound of the "weed" in a net bag and boihng in a gallon of water. The water on cooling sets to a jelly. The calves are given one glass of jelly in their milk each meal and wonderful results are said to be obtained." The high proportion of mineral matter is noteworthy ;i but without making a fuller investigation, it is impossible to say precisely wherein lies the value of this seaweed. Purple laver {Porphyra laciniata), a source of Japanese amanori, is found in abundance on the rocky shores of America and Europe generally; but it is not used in this country save sparingly by the Chi- nese, who usually import it directly from China, and by some of the Indians of our northwest coast. In Ireland it is known as 'sloak,' and is boiled and served with butter, pepper, and vinegar as an ac- companiment of cold meats, or is served with leeks and onions. Dulse (Rhodymenia palmata) is found abundantly on rocky shores both in this country and in Ireland. It is very abundant in New England, where it is rough-dried in the sun and eaten as a relish. In Philadelphia it is caUed sea-kale and eaten as a vegetable. In Scot- land it has long been used both in the fresh state and dried. In the Scotch Highlands, "a dish of dulse boiled in milk is," it is said, "the best of all vegetables." In Ireland, it is eaten with fish or boiled in milk with rye flour. Purple dulse {Iridea edulis), which occurs on the Pacific coast, is often eaten like Rhodymenia palmata. iCf. Analysis of Chondrus crispus, p. 254. Nutrition Investigations. 257 Besides such lichens and algae, and the bark of trees, various tubers are used as food for man. In Japan, the tubers of Hydrosme rivieri (ConophaUus Konjaku) are extracted with lime water, and the result- ing gelatinous mass is cut into small cakes. These, cooked with "shoyu" or Soy-bean sauce form a common article of diet. The tubers of many species of Orchis and Eulophia, native to Turkey, the Caucasus, Asia Minor and the greater part of Central and Southern Europe, furnish a food material known as Salep. The small ovoid, oblong or palmate tubers are decorticated, washed, heated till horny and semi-transparent, and finally dried. An abundant mucilaginous extract is obtained by macerating the bulbs in water. Frequently the tubers are ground to powder, and the powder used like sago or tapioca. Royal salep, said to be used as food in Afghanistan, is pre- pared from Allium Macleanii. A former instructor in the American College for Girls, in Constantinople, reports that salep is a very com- mon article of diet in Turkey. It is sold in the markets in powdered form, and is made into a sort of sweetened gruel with milk. Not only is it used as a warm drink in the household, much as we use cocoa or chocolate, but it is also sold in the streets by venders, who either stand in booths along the way, or go about carrying huge brass urns strapped to their shoulders, clinking their cups and calling " Taze- Sahlep!"! It is especially popular in districts of the city where peo- ple work late at night. In the month of Ramazon, the time of all-day fasting, hot salep finds a ready sale at night. It is no uncommon thing to see the workman standing with his salep cup in hand, waiting for the firing of the sunset cannon. In spite of the fact that there have been almost no scientific inves- tigations as to the digestibiUty of such mucilaginous plant substances there seems to be a special virtue attached to mucilages in the popular mind. The prevailing impression is shown in some of the following remarkable statements. The United States Dispensatory, 1908, not only says that the mucilaginous extract of slippery elm hsirk {Ulmus fulva, Michaux) is nutritious, but adds, "We are told that it has proved sufficient for the support of life in the absence of other food." Of salep Smith (25) says in his dictionary of economic plants: "It con- tains a chemical substance called bassorin, which is said to contain more nutritious matter than any other vegetable product, one ounce per diem being sufficient to sustain a man"! The United States Dis- pensatory also assures us that salep is "highly nutritious." Johnson ^Fresh salep. 258 Mary Davies Swartz. (7) particularly recommends Iceland moss {Cetraria islandica) as a diet for consumptives, as "it seems to be both extremely nutritious and very easy of digestion, though of course, only capable of use as a substitute for starchy matters." In regard to Irish moss {Chondrus crispus), he is a little more uncertain. "It is much used for invalids, especially in cases of consumption, but with doubtful advantage when substituted for more nutritious food." Schneider (21) says of Ice- land moss: "Inhabitants of Iceland, Norway, and Sweden mixed this lichen with various cereals and mashed potatoes, from which an un- commonly healthful bread was prepared." Until the matter has been thoroughly investigated, we must suspend our judgment as to the ac- curacy of such statements. After a few metabolism experiments, Oshima (15) far more conservatively remarks concerning the algae of Japan: "Their actual value doubtless depends in considerable measure upon the mineral salts they contain." In view of the scarcity of any scientific investigations as to the be- havior of all these substances in the body, further experiments upon their nature and digestibility seem highly desirable, since they are not only widely distributed, and already form a considerable portion of the diet of many persons; but because, if they possess any real nutri- tive value, a wider use of such comparatively cheap materials would be an economic advantage; and because, under the prevailing notions as to their food value, they are sometimes relied upon as a source of nutriment in diseases (as diabetes) where the character of the diet is particularly important. The present work has been undertaken to throw some Hght on this interesting subject. A survey of the litera- ture shows that even the chemical nature of many of these algae has scarcely been investigated; and if this were known, we should still be under the necessity of studying their behavior in the animal body, for it is impossible to tell from chemical analysis alone whether a given substance will or will not prove digestible, as Rubner has long since warned us. II. HISTORICAL PART. Introduction. According to the current practice of agricultural analysts, the car- bohydrates of plants are reported as crude fiber and nitrogen-free extract. Crude fiber is the term applied to the resistant mixture form- ing the mature cell wall, shown as long ago as 1864 by Henneberg and Stohman (41) to have no definite chemical composition. It is there- fore not identical with cellulose, but consists of a mixture of cellulose with incrusting substances, lignin and cutin, the relative proportions of which have recently been exhaustively studied by Konig (51), Fiirstenberg (39), and Murdfield (63). Cellulose is the chief consti- tuent; the other two are usually present in varying proportions. Schulze (74) to whom much of our knowledge of the composition of the plant cell wall is due, has classified the carbohydrates of the nitro- gen-free extract as follows : I. Water-soluble carbohydrates. To this class belong the mono-, di-, and tri-saccharides, and some soluble polysaccharides. II. Carbohydrates insoluble in water, but yielding sugar under the action of diastase. The chief member of this group is starch. III. Carbohydrates insoluble in water and resistant to the action of diastase, never being changed by it into sugar. This group is called the Hemicelluloses. The term hemicellulose, as used by recent writers^ seems to be inter- preted to include some polysaccharides of the first group. It is there- fore used here as a group name for those carbohydrates which are dis- tinguished from cellulose by being capable of hydrolysis on boiling with dilute mineral acids, and from the other polysaccharide carbohy- drates by not being readily digested by diastase. According to the kind of sugar yielded on hydrolysis, the hemiceUuloses are designated as Pentosans or Hexosans, the latter including Galactans, Mannans, Dextrans, Levulans, etc. After a general review of the chemical ^e.g., Lohrisch. 259 260 Mary Davies Swartz, nature of lichens and algae, each of these classes will be discussed separatel);" in detail. The percentage composition of some common species of algae is shown in the following table : WATER. PROTEIN. FAT. CARBOHYDRATES. FOOD MATERIAL. ^'fref°' Crude Extract. ASH. I.* Cystophj'llum fusiform, dried 15.74 18.75 13.53 23.08 13.98 18.92 80.00 80.00 80.00 13.40 11.37 9.58 19.35 7.11 33.75 11.61 1.4 3.7 1.8 13.06 0.32 .49 .46 1.73 .87 1.30 .31 0.0 0.0 0.0 2.59 1.2 54.84 51.63 9.79 46.18 47.70 41.22 37.81 14.4 12.5 14.1 54.16 2.57 43.3 5.3 17.56 Ecklonia bicyclis, dried. . Enteromorpha linza, dried (Limu eleele).. . Laminaria sp., dried.. . . Porphyra laciniata, dried 9.79 19.21 21.24 9.75 Ulopteryx pirmatifida, dried 31.35 Il.t Ahnfeldtia concinna, fresh (Limu akiaki).. . Ulva fasciata and U. lactuca, fresh (Limu Dahapaha) 4.2 3.8 Gracilaria coronopifolia, fresh (Limu manauea) IILJ Chondrus Crispus, dried. IV. § Cetraria islafidica 4.1 14.2211 2.2 ♦Oshiina (15). t Reed (18), (calculated on uniform water basis). tAnnet (1). § Schmidt (24), first studied the ash and reported a notable amount of calcium and potassium phosphates. He found no nitrogen. Blondeau (3) reported 21.36 per cent nitrogen. II Brown (334). Until 1905 the chemical nature of the constituents of algae had received little attention. Analyses of many species of algae from Japan and China were reported recently by Konig and Bettels (8), the results of which are given in the following table on page 255. According to Oshima and Tollens (16) the carbohydrates of Por- phyra laciniata consist largely of anhydrides of J-mannose and i-ga- lactose. Miither and Tollens (13) studying various species of Fucus (F. vesiculosus, F. nododus, F. serratus), Laminaria, and Chondrus crispus, found a methyl-pentosan (fucosan), in Fucus and Laminaria; and glucose, fructose, galactose and pentose groups in Chondrus Krefting reports a reserve carbohydrate in Laminaria digitata in win- Nutrition Investigations. 261 >-, ^^S >> 43 X! aj -W ooououotjuo y 333idS3:Sp3 3 fcnVjUlHl-'l-i'-i'-i'Hti SH O O o c o o o ty (D d) 1) O tn tn ^ c/a c/1 O O O O ■ O ^_) J_! +-> +J *J O U CJ CJ O 03 rt _rt _rt ^ C^ rt C^ rt rf o o o o o O aj ^ rt jrphyra . . arphyra te elidium R elidium B elidium ca aminaria ther Lami o a. a o o o d; ph o o o hj o O O W 3 -2 CqoO-*iO«3t^COC3g 262 Mary Davies Stvartz, ter only, which yields J-glucose. The algae investigated are thus all seen to yield pentoses, very frequently fructose and methyl-pentose, sometimes glucose and galactose. Lichens are symbiotic forms embracing algae and fungi. Because of this symbiotic nature, they exhibit great variety in composition. From the investigations of Escombe (6), Ulander and Tollens (27), Karl Miiller (11), Nilson (14), Wisselingh (29) and others,i it appears that the cell walls are usually of cellulose, but occasionally of chitin.^ Many species yield on extraction with hot water a gelatuiizing sub- stance, which Berzelius (2) in 1808 named " Flechtenstarke " {lichenin), but which later investigators^ have shown to be, not a single substance, but a number of related carbohydrates yielding dextrose, such as lichenin from Cetraria and Ramalina fraxinea, and evernin from Ever- nia prunastre, usnin from Usnea barbata. Other species, on the con- trary yield little dextran, but mannan, galactan, pentosan and methyl- pentosan in varying proportions. The table on page 257 showing the hemi-celluloses occurring in a number of lichens, has been compiled from data given by Karl Miiller (11) and Ulander and Tollens (27). Occurrence and Nature or Cellulose. Cellulose is said to occur in pure form in the wall of the young plant cell. With increasing age, modifications take place by which the true cellulose becomes more and more encrusted with Hgnin and cutin, two substances shown by Konig (52), Fiirstenberg (39), and Murdfield (63) to be almost entirely indigestible. According to Wielen (87) and Hofmeister (43), even pure cellulose is not a simple substance, but can be separated into soluble and insoluble portions.'* Much of our in- formation regarding the nature of cellulose is due to the work of Schulze and his pupils. Schulze (75) has defined cellulose as that part of the cell wall giving the typical cellulose reactions,^ and yielding dextrose on hydrolysis with concentrated sulphuric acid. Tor early literature see Czapek, Biochemie der Pflanzen, Vol. I, pp. 514-516. -Chitin occurs in Peltigera canina and Evemia prunastre. ^Cf. MuUer (11) and Ulander (26). ^According to its beha\'ior in sodium hydroxide solutions, the quantitative rela- tions depending upon the source of the cellulose and the concentration of the solu- tion. ^InsolubiUty in dilute acids and alkahes; solubility in ammoniacal copper oxide solutions; and production of a blue color with iodine and sulphuric acid. Nutrition Investigations. 263 CO en CO c/2 c/i en (U O (U .^ -J-l -M 4-> d d c! d c: 1:1 Gj !D (D CD . T^ 'P. '>.'>. '>i •5 5 :S -5 -5 :S -u is -u 5i fi ^ c >^ >^ >^ >> XI ^ ^ ^ 4-> +J +J +J 4-> O O O O4J-M+J-M . -T" oj aj (u D oj « ><1>< +J -M .*-> O O U CJ ci 03 cd c3 c^ c3 g3 g3 0000 '13 "TS "^ -13 en tn en 000 o w S^ G c^ _2 TT, 'S "J K a "S -s .2 =^ o ° S g O CJ O H i-J ^ s I S o 2 o 2 M El ID hJ O a, O ,J-I J rt KJ ^ 8 .£? g O ^ 0) tn PL, CA! fL| P O T-nfMCO^iOcOt^COOOi-iiM 264 Mary Davies Sivartz, CYTASES IN THE VEGETABLE KINGDOM. By the early investigators, Haubner (40), Henneberg and Stohman (41), Kuhn,Aronstein,and Schulze (54), it was accepted without much question that, since cellulose disappeared from the alimentary tract of herbivora, it is digested like starch, and equally valuable as a nu- trient. But after Tappeiner (78), in 1884, showed that cellulose could be decomposed by micro-organisms, and promulgated his theory that this was the only way to account for the disappearance of cellulose from the alimentary canal of ruminants, the matter fell into great dis- pute,' and the question is not yet definitely settled as to how cellulose is digested and what are the products of its digestion. A diligent search has been made for enzymes capable of attacking it {cytases), but so far, such cytases have been proved to exist only in plants and lower animals. Many of these so-called cytases act upon hemicel- lulose rather than true cellulose, and will be discussed in connection with the hemicelluloses, though it is not always possible to make a sharp distinction between the two. A careful review of the subject of cytases in plant physiology up to 1898, has been made by Bieder- mann and Moritz (34), from which it appears that the penetration of wood by the mycelia of moulds is due to such cytases, and that a powerful cellulose-dissolving enzyme has been derived from Peziza sclerotium by de Bary (37) and from another botrytis (presumably a Peziza) by Ward (84), while Brown and Morris (36) have described cytases existent in germinating grasses which dissolve their cell walls. That this is anything more than a diastatic enzyme is denied by Rei- nitzer (67) ; but Newcombe (64) considers the assumption of the iden- tity of all cell-wall dissolving enzymes with diastase as far from jus- tifiable. Bergmann (32) 'reports such cytases in hay and straw. Scheunert and Grimmer (71), on the contrary, find none in oats, corn, horse-beans, lupine seeds, buckwheat or vetch. Thus we see that even in the case of plants, these enzymes need to be isolated and identified before we can arrive at any satisfactory conclusions. That cellulose can be dissolved by bacteria has been demonstrated for such forms as Amylobacter butyricus, Vibrio regula and Clostridium polymyxa (34). Omelianski (65) has described two organisms which ferment cellulose, and Ankersmit (31) finding Omelianski's bacteria on hay, has studied their behavior when introduced into the alimen- tary canal of the cow on its food. He finds that they do not increase ^For a review of this discussion cf. Lohrisch (56). Nutrition Investigations. 265 in number during their passage through the digestive tract, and there- fore concludes that they play a very inconsiderable role in the decom- position of cellulose. According to Van Iterson (81), certain aerobic bacteria, attacking cellulose, form from it products which nourish other forms {spirilla); certain anaerobes are also shown to attack it. Eberlein (38), finding in the first stomach of herbivora Infusoria which utilize cellulose for food, suggests that these protozoa, digested farther along in the ahmentary tract, serve as means of transforma- tion of cellulose into products which the animal can digest; but there is nothing to indicate that such forms occur in sufi&cient numbers to be worthy of much consideration. Since 1906 three investigators have given the problem careful at- tention. Scheunert (68) has concluded from experiments in vitro that bacteria play an exclusive role in the solution of crude fiber in the coe- cal contents of horses, swine, and rabbits. He found that filtered coecal fluid acted on cellulose much less than imfiltered or simply strained coecal contents. This is contrary to the opinion of Hof- meister (45) and Holdefleiss (48), who attribute the phenomenon to the action of enzymes, and explain the loss of power occasioned by filtering as due to the effect of exposure to the air upon the enzymes. Lohrisch (57) has reported that fresh coecal fluid is efi"ective in destroy- ing cellulose while heated fluid is not. On the other hand, implanting the sterilized fluid with coecal bacteria and protozoa would not restore its activity. Coecal flmd kept at 38° C. any length of time gradually lost its ceUulose-dissolving power, while that kept on ice remained active, v. Hoesslin and Lesser (47) have attempted to explain these apparent contradictions, and conclude from their own experiments that anaerobic bacteria are the most effective agents in cellulose de- composition in the intestine. Equal volumes of non-sterilized and sterilized coecal fluid of the horse, to which weighed amounts of cel- lulose had been added, were suspended in sterile physiological salt solu- tion under practically anaerobic conditions and digested for periods of from 9 to 35 days. The disappearance of cellulose with the non-steril- ized coecal fluid amounted to from 55.7 per cent to 71.2 per cent; with sterflized fluid, to from 6.2 per cent to 42.4 per cent. It was also found that the addition of 1-5 grams of dextrose would effectively protect the cellulose from digestion by the non-sterflized fluid, the bacteria preferring the more easily attacked carbohydrate. The gases evolved in these fermentations were characteristic of bacterial action, being chiefly methane, carbon dioxide, and hydrogen. The retarding effect of exposure to the air is explained by the theory that anaerobes are 266 Mary Dames Swartz, the effective agents. So, also, the fact that Lohrisch was unable to get cellulose digestion in sterilized fluid again inoculated with unsteril- ized fluid is attributed to the mediiun's being an unfavorable one for the development of these organisms, inasmuch as the addition of pep- tones to similar preparations caused in several cases an increased de- composition. It seems fairly well estabhshed, therefore, that the action of the coecal fluid of the horse is due to enzymes of bacterial origin. CYTASES IN LOWER ANIMALS. There is no doubt that cytases occur in some of the lower forms of animal life. Biedermann and Moritz (34) fovmd a powerful cellulase in the secretion of the Hver of the common snail {Helix pomatia), and their observation was verified by E. Miiller (61), also by Lohrisch (57) who reports two series of experiments in which snails fed tender let- tuce leaves digested from 40.1 per cent to 81.6 per cent of the cellulose present. On the other hand, Miiller (61) could not verify Knauthe's report of a cellulase in the hepato-pancreas of the carp (50) ; Pacault f otmd none in the saUva of Helix pomatia (66) ; and Biedermann none in the digestive juice of the meal worm {Tenebrio molitor) or of the cabbage worm {Fieris brassica) (34) . Biedermann also examined the faeces of the cabbage worm microscopically and foimd unaltered par- ticles of leaves, from which he concluded that much of the plant food eaten is excreted unchanged. Lohrisch (56) has obtained similar re- sults with caterpillars of sphinx moths {Sphinx euphorbiae), not only in experiments with intestinal juice in vitro, but also in feeding expe- riments in which the cellulose was quantitively excreted. SeUiere (75-76) has recently added some interesting contributions to this subject, showing that cotton treated in various ways; namely, that recovered after solution in Schweitzer's reagent, that treated wdth concentrated zinc chloride, or with 25 per cent caustic alkali hot or cold imtil the fibers are swollen, and subsequently washed with 1 per cent acetic acid and water, is attacked by Helix pomatia much more readily than the untreated substance. Subsequent drying of the treated cotton diminished its digestibihty somewhat, suggesting that the physical condition of the cellulose is a definite factor in its utilization. Selliere believes that only the more tender portions of plant cellulose are attacked by the digestive juice of this snail. It would seem that the previous treatment of ths cellulose is a factor to be kept in mind in the interpretation of the results of feeding experiments.^ ^Cf. the experiments on cellulose utilization in the dog, p. 263. Nutrition Investigations. 267 CYTASES IN HIGHER ANIMALS. There is at present no proof of the existence of cytases in any of the higher animals. The literature on the subject has been exhaustively reviewed by Bergmann (32), and Lohrisch (55, 56, 57) audit appears that there is no cellulase in the saliva or pancreatic juice of swine, horses, cattle, or sheep. The old observation by MacGillawry^ (cited by Biedermann and Moritz(34) that a cytase can be extracted from the vermiform appendix of the rabbit has been denied by Zuntz and Degtiareff (88). Schmulewitsch's- statements (also cited by Bieder- mann and Moritz) are worthless because he employed no antiseptics. E. Miiller (61) foimd no sugar formed from the decomposition of cel- lulose in the stomach of the goat, and Lusk (59) observed no increase in sugar elimination after feeding a phlorhizinized dog 20 grams of cauUflower, or a phlorhizinized goat 10 grams of paper. Lohrisch (57) fed pure cellulose (5-20 grams) to a phlorhizinized rabbit and fovmd that it had no marked influence on the sugar output, and no nitrogen- sparing effect. Scheunert (70) has made further investigation on the action of the saliva and salivary glands in sheep, and confirms the earlier experiments with the saliva of this animal. On the other hand, Selhere (77) reports that the specially treated cellulose mentioned above is converted into dextrose by the intestinal secretions of the guinea pig in some instances. Practically nothing is known concerning the way in which cellulose disappears from the aHmentary tract of man. Schmidt and Loh- risch (73) fed pure cellulose to diabetics and observed a disappearance averaging 77.7 per cent, and no increase in the elimination of sugar. They believe that most of it is absorbed in soluble form and not de- stroyed by fermentation in the intestines. Lohrisch, having fed cel- lulose in various diseases of the alimentary tract,^ calls attention to the fact that in constipation, where there is the least bacterial action, the utilization of cellulose is highest, while in fermentation dyspepsia, in which one might expect a marked disappearance, the utilization is lowest. He therefore considers the digestion of cellulose as due at least in part to enzymes. ^Archiv Neerland, Vol. XI. ^tJber das Verhalten der Verdauungssafte zur Rohfaser der Nahrungsmittel. Biilletin de rAcademie Imperial de St. Petersburg, 1879. ^See results, p. 264. 268 Mary Davies Swartz, Digestion and Utilization of Cellulose by Animals. The literature on the digestion of cellulose up to 1909 has been so exhaustively reviewed by Lohrisch that it is unnecessary to enter into a detailed discussion of it. From tables (55) showing the results of all previous experiments on the utiUzation of crude fiber in herbivora, carnivora, and birds, it appears that in the case of herbivora, especi- ally ruminants, 20-28 per cent of the crude fiber ingested with food disappears from the alimentary canal; that in case of carnivora^ and birds^ there is no utilization whatever. Lohrisch (56) himself reported three experiments in which dogs were fed pure cellulose and digested 31.1 per cent, 37.45 per cent and 5.4 per cent respectively, but Scheu- nert and L5tsch (72) repeating Lohrisch's work with a somewhat dif- ferent method of determining cellulose found that the administration of 40 grams of prepared white cabbage, containing 7.37 grams of pure cellulose, resulted in the recovery of the total amount ingested. Cook- ing the cabbage in bouillon did not increase its digestibility. They attribute the apparent utilization in the preceding experiment to des- truction of cellulose by the reagents used for its purification. Since the publication of their paper, Lohrisch has repeated his work with the dog (57), and reports complete recovery of the cellulose fed. He explains the error in the earlier investigation as due to the fact that the ingested cellulose was twice subjected to purification (before feed- ing and in faeces) with consequent increase in percentage of loss, which was not taken into account. He points out the inevitable loss of some cellulose by any method at present in use for its determina- tion, and defends his own as sufficiently accurate for all practical pur- poses if conditions are carefully observed.^ ^The only experiments on record are by Voit and Hoffmann on the dog and by von Knieriem on the hen. ^Experiments by Weiske on the goose, and by von Knieriem on the hen. 'Lohrisch used the method of Simon and Lohrisch, in which the cellulose is dis- solved by heating for an hour on a water bath with 50 per cent potassium hydroxide, then adding f cc. of 30 per cent hydrogen peroxide, and digesting from § to | hour longer if necessary. The cellulose is then precipitated by adding to the solution one half its volume of 96 per cent alcohol and 6-7 cc. of concentrated acetic acid; filtered off, washed with water, dilute acetic acid, alcohol and ether, dried and weighed. Scheimert and Lotsch mix the substance to be analyzed with 100 cc. of cold water, add 100 grams of potassium hydroxide and heat for an hour on a water bath, then filter through a hard filter paper, wash the residue on the paper with boiling water till only a trace of alkali remains, transfer it to a beaker and thence to a weighed Nutrition Investigations. 269 Cellulose digestion in the dog has been almost simultaneously stud- ied by V. Hoesslin (46). Two dogs on a meat-fat diet to which was added daily 2 grams of specially prepared white cabbage (containing 63.25 per cent of pure cellulose), for five periods of five days each, excreted on the average 99.7 per cent and 94.5 per cent respectively. This long experiment is significant as showing no adaptation of the digestive glands to the type of food. By these independent workers it seems now well established that the dog is unable to utilize cellulose. Hoft'mann (42) has just published the results of some investigations on the influence of cellulose on the nitrogen balance and on phlo- rhizin-diabetes in the rabbit, from which it appears that after inges- tion there is no increase of sugar excretion, and no glycogen formation, yet he thinks that cellulose and hemicelluloses have a favorable influ- ence in phlorhizin-diabetes.^ It seems to follow from this, that even in case of herbivora cellulose is not utilized in the manner customary for starch and sugar. DIGESTION AND UTILIZATION OF CELLULOSE BY MAN. A similar tabulation of results of feeding experiments on man, shows that cellulose is not so well utilized as by herbivora, but does disap- pear in appreciable amounts. With one exception, the cellulose in all these experiments was administered as crude fiber. Hofmeister (43) fed pure cellulose and reported 75.7 per cent soluble cellulose and 5.6 per cent insoluble cellulose digested. Konig and Reinhardt (53) added to a diet rich in protein and fat, but free from cellulose, in sev- eral experiments, green peas and ripe shelled peas, red cabbage, white filter, on which it is washed successively with hot water, dilute acetic acid, hot water, alcohol and ether, and finally weighed. Scheunert and Lotsch claim that by Lohrisch's method the cellulose is altered in character, and as much as 40 per cent lost in the process; and that subsequent treat- ment of the recovered material causes an even greater per cent of loss, while by their method the loss in the first case is not over 6.8 per cent, and that in the second case even less. For the details of this controversy over method see the following : Simon and Loh- risch; Zeitschrift fiir physiologische Chemie, Vol. 42, p. 55, (1904). Scheunert; Berliner tierarztHche Wochenschrift, No. 47, p. 826, (1909) . Scheunert and Lotsch; Ihid., p. 867, (1909); also Biochemische Zeitschrift, Vol. 20, p. 10, (1909); and Zeitschrift fiir physiologische Chemie, Vol. 65, p. 219, (1910). Scheunert and Grimmer; Berliner tierarztHche Wochenschrift, No. 48, p. 152, (1910). Lohrisch; Zeitschrift fiir physiologische Chemie, Vol. 69, p. 143, (1910). ^Unfortunately the original paper was not accessible. 270 Mary Davies Sivartz, beans, graham and soldiers' bread and found 30.27 per cent to 76.79 per cent of the added cellulose digested. Lohrisch (55) finds that the cellulose of a common vegetable diet disappears from the alimentary tract in large amounts, the actual quantity varying with the age, source and tenderness of the cellulose. Thus he finds that for normal indi\dduals, of cellulose from lentils, 45 per cent is digestible; from kohlrabi, 79.1 per cent; from white cabbage, 100 per cent. Under abnormal conditions in the digestive tract, he has obtained the fol- lowing results: CONDITION. CELLULOSE UTILIZATION IN PER CENT. Normal 57.9 Chronic Constipation 81.4 Fermentation Dyspepsia 37.8 Gastrogenic Diarrhea 29.5 Fatty Faeces in Icterus 27.8 Fatty Faeces in Disease of Pancreas 20.9 According to Lohrisch, two diabetics on a cellulose-free diet, to which white cabbage was added in quantities to yield about 6 per cent of cellulose per day, digested 68.6 per cent and 84.5 per cent respectively, without increased output of sugar in the urine. Since the only way to determine definitely the energy value to the organism of such amounts of cellulose as are absorbed, is by means of respiration experiments, Lohrisch (57) has performed such an expe- riment on man, using the Zuntz-Geppert apparatus. In fasting, the respiratory quotient averages about 0.76. After ingestion of carbo- hydrates such as starch, it rises gradually in two to three hours, to 0.9-1.0, and when the carbohydrate has been consumed, sinks again to a lower level. Since the respiratory quotient for fat is 0.7 and for protein about 0.8, it is possible to determine in this way to what extent the carbohydrate replaces protein and fat in metabolism. Hence if cellulose is absorbed and oxidized as a carbohydrate, the res- piratory quotient should rise. If it is decomposed by bacteria, the respiratory quotient should not rise, since the theoretical respiratory quotient for fatty acids, such as butyric and acetic, is, according to Mimk (62) and Mallevre (60), 0.6 and 0.5 respectively. Now Loh- risch, feeding a man moist cellulose equivalent to 73.6 grams of dry substance, of which 25 per cent was digested (18.5 grams) obtained the following results: Nutrition Investigations. 271 ■NOII -saoNi asoTxiTiao a CM ^ lO CO t^ 05 tH CM 1-1 1-1 KO ONiMNioaa «aiav tS ^ i-llN rtlN H|« iHiN •NOixonaoHd jod 1 1 1-1 CM CO T|H CO lO 1 1 1 + 1 + HN s -^l^ rHiN CO O •NoixamnsNoo to 1—1 + CM 1 lO CO CO I— 1 ^H Ttl ++++++ -* o (M OS CO CO 1-H CO 00 CO CO •axmuK lO ^ l^ 00 1-1 00 CO t^ CM CM Haa NOiioaaoHj soo o ,— 1 £2 E2 O 1> 1— 1 t^ CO 05 1—1 '^ Ui 1—1 1— ( lO ^ ■* lO -# lO 1— 1 1— ( 1— 1 T-H 1— ( T-( I> o CO r^ tn Oi Ci CO CM I> CO •aiaNiK 4''^ 00 '-' -;>< CO CO ^ CO 00 CO lO Haa NOLMIMSNOO zo § ^ Ci CO -* r^ CM CO CO CO 1-1 05 O O O CM i-l 1— 1 1—1 I— 1 CM CM CM CM CO CM Tt* - t^ t^ t^ CO 1> o o o o o o o o o o o Ico on r^ o CO O CM O ^ 1-1 •Noiionaona zoo "^ CO o CO ^ CM 1-1 CM T-i O O feco CO CO CO CO CO CO 00 CO CO S TjH r^ o CO CO t^ CO ■^ lO lo •NOixdimsNOO 200 8 "-I t^ lO '^ CO CM CO -^ CO 1-1 rt< "* ^ Tt< tH ^ TiH -* 'cH *; S 1^ o Ci 00 O 1-1 Oi o S q; Tt< CO O ^ »0 03 aaivHNi aKaioA S 52 Ci 2i CM CO t^ CO O CO CM ■^ rtl -^ lO ■>* lO •saia Tt< r- CO t^ • CO Tf* lO CO ir^ CO C3 O -laaaxa ao aaawmi 272 Mary Davies Swartz, The respiratory quotient attains its highest value in the fourth hour, instead of the second or third, showing that cellulose is absorbed more slowly than starch. The rise is too slight to indicate that cellu- lose exercises any considerable protein- or fat-sparing eSect. It is unfortunate that the amount of cellulose absorbed was so small. It is striking that the 02-consumption decreases at the very time that the respiratory quotient rises, and the COo-production scarcely in- creases. Lohrisch interprets this as indicating that the increased 02- consumption required for oxidation of the cellulose is compensated by a sparing of protein and fat. The difierences seem too small to draw any satisfactory conclusions as to the energy value of cellulose. The low respiratory quotient in the later hours of the experiment, together vAxh the increased 02-consumption, indicates the utilization of some of the cellulose in the form of fatty acids. We must bear in mind that no formation of sugar or glycogen from cellulose, in men or ani- mals, has been demonstrated. Further investigations woiild seem to be necessary before we can agree with Lohrisch in saying, " Wir wissen, dass Cellulose und Hemi-cellulosen vom Menschen reichlich verdaut werden, wir haben alien Grund anzunehmen, dass ihre Verdauung nach Analogieder Starke ahlduf I . . . Die resorhirtenM engen werden im menschlichen Organismus vollstdndig verbrannt. Dabei wird Eiweiss und Fett von der Verbrennung geschiitzt.^' In any event, the quanti- ties of celliilose which the alimentary tract of man is capable of ab- sorbing are, apparently, too small for it to play a role of any impor- tance in the diet of a normal individual. OCCUREENCE AND NATURE OF PENTOSANS. The anhydrides of the 5-carbon sugars are collectively designated as pentosans. These are not reported to occur in the animal kingdom, but the pentose sugars are found forming a part of the nucleic acid radical of the nucleo-protein molecule. In the vegetable kingdom, pentosans are very widely distributed, as has been shown by many investigators, especially Tollens and his pupils. ^ They occur in all kinds of plants, from the lowest to the highest, and are limited to no iToUens, Landw. Vers., V. 39, p. 401, (1891); Tollens, Jour. f. Landw., Vol. 44, p. 171 (1896). For an exhaustive review of the literature on the occurrence of the pentosans see V. Lippmann, Chemie der Zuckerarten, 3rd Edition, Vol. I, pp. 44-60; 116-123; and Czapek, Biochemie der Pflanzen, Vol. I, pp. 537-545 (1905). Nutrition Investigations. 273 particular organ or tissue, being found abundantly in roots, stems, leaves or seeds. In regard to solubility in water, pentosans show all possible varia- tions. De Chalmot (108) found them present in the watery extract of the leaves of many plants; Winterstein (167) in the somewhat mucila- ginous hot water extract of the seeds of Tropaeolum majus; Schulze (146) , in both soluble and insoluble form in the cotyledons and endosperms of the seeds of Lupinus luteus and other legumes, where they are doubt- less stored as reserve material for the growing plant; and in the cell walls of the mature plants, where in most cases they approach true cel- lulose in character. It is difficult to differentiate these highly resis- tant pentosans of the cell wall, which are commonly included in the term crude fiber, from the hgno-celluloses and oxycelluloses also found there, which as Cross, Bevan and Beadle (104) have shown,i are like true pentosans in yielding furfurol on distillation with dilute hydrochloric acid. Besides hemicelluloses yielding pentoses {xylose and arabinose) exclusively, occur many 3delding also methyl-pentoses (fucose, rhamnose) . These yield on distillation with dilute hydrochloric acid, methyl-furfurol, which is precipitated by phloroglucin, and hence included in quantitative estimations of pentosans by the method of Tollens andKrober (121). The distribution of methyl-pentosans has been studied especially by Tollens and his pupils. Japanese "Nori" {Porphyra laciniata, Laminaria, and other seaweeds) (129), tragacanth and many other gums (163) contain fucosan. Rhamnose occurs also widely distributed in the plant kingdom, but more frequently in the form of a glucoside. Rohmann (134) reports a rhamnosan in Ulva lactuca. It is a very common thing to find pentosans and hexosans occurring together. In fact, it is absolutely impossible, in treating of hemicellu- loses, to draw any sharp dividing lines, for they are not only intimately associated, but frequently chemically combined. Schulze (146) has given the name paragalactan to the carbohydrate yielding arabinose and galactose, which occurs in the seeds of many legumes. Winter- stein (167) finds galacto-xylan in the water extract of Tropaeolum majus, and numerous other examples of such combinations might be cited. A class of substances to which has been given a distinctive name because of their peculiar gelatinizing property, is the Pectins. As Czapek^ remarks, "It is uncertain whether they form a definite iFor further details see v. Lippmann; Chemie der Zuckerarten, Vol. I, pp. 160-169. ^Die Pektin-Substanzen; Czapek, Biochemie der Pflanzen, Vol. I, p. 545. 274 Mary Davies Swartz, class of cell wall substances, or whether they should be classified as 'hemicelluloses' or 'pentosans.' " In 1868, Scheibler (141) found a sugar which he called pectinose, but which was later shown to be ara- binose (142), In 1875, Reichardt (132) obtained a pectin body from carrots and beets, which he called 'pararabin,' expressing the view that pectins should hardly be considered as a special class of carbo- hydrates. Tromp de Haas and ToUens (160) have found from numer- ous analyses, that the pectins do not differ from other carbohydrates in their relative proportions of hydrogen and oxygen so much as earlier workers supposed, and hence they may be classified with other hemi- celMoses according to the products of their hydrolysis (pentoses; galactose and other hexoses). Cross (106) believes them to be allied to the ligno-celluloses. The whole matter is still in a state of uncer- tainty. Herzfeld (116) has shown that arabinose can be obtained from most pectins, and consequently they have been included among the pentosans, though from the frequency with which they yield ga- lactose, they might equally well be discussed with the galactans. Ac- cording to Czapek while pectins occur frequently in phanerogams, ferns and mosses, their presence in algae is doubtful, although it is possible that soluble carbohydrates of algae 5delding arabinose or ga- lactose are closely related to the pectins of other plants. ^ Role of the Pentosans in Plant Physiology. Comparatively little is known of the role of pentosans in plant phys- iology. De Chalmot's (108) observation that they decrease in quan- tity in seeds — peas and corn — during germination, and reappear in the stems and roots of the growing plant, would seem to indicate that they form a part of the reserve material in the seed; but Schone and Tollens (145), finding no diminution in the amo\mt of pentosans in grains during germination, but rather a shght increase, declare that they do not belong to the reserve-stuff of the seed; so the question may be regarded as still unsettled. Changes in the relative amounts of pentosan in plants at different stages of growth, studied by Cross, Bevan and Smith (105), Gotze and Pfeiffer (113), Calabresi (98), and others, show that the increase of pentosans runs parallel to the forma- tion of the skeletal substance ; and have led to the idea that they arise through the transformation of a part of the cellulose, and along with lignin and cutin, take part in wood formation. Ravenna and Cereser ^Cf. also Bigelow, Gore, and Howard (92). Ntitrition Investigations. 275 (131) find in the case of dwarf beans that when the food is wholly dex- trose administered to the leaves, pentosans increase greatly, especially in the light, and that when the functioning of chlorophyll is prevented for long periods the amount of pentosans decreases. They conclude that the simple sugars exert a preponderating influence in pentosan formation, and that these serve as a reserve material when the plant has exhausted its more readily available food materials. PENTOSANASES IN THE VEGETABLE KINGDOM. Our knowledge of enzymes inverting pentosans is meager, and rather indefinite. The action of such forms as Hymenomycetes upon wood seems to be of chemical nature. At any rate it is evident (107-146) that they are able to utilize xylan. Bourquelot and Herissey (95) have isolated an enzyme from malt diastase which produces reducing sugar from pectins, and call it pectinase. This is not to be confused with the so-called pectase which causes the coagulation of pectin bodies. Bigelow, Gore and Howard (92) also find that the enzymes of Asper- gillus partially hydrolyze the pectin of gentian root. According to Harrison (114), Bacillus (?/grace« produces a cytase capable of dissolv- ing the cell walls of potatoes, turnips, cauliflower and allied plants, which acts particularly on the middle lamella, the supposed seat of pectin.i The latter is not an inverting enzyme. In Persian Berries (Rhamnus) (162), in Penicillium glaucum, and Botrytis cinerea (90), an enzyme (rhamnase) has been foimd which splits ofl rhamnose from some of its glucosides {rhamnetin and rhamnazin). An early observa- tion of the presence of rhamnase in the rutin of garden rue was made by Borntrager (94). That some of the so-called cytases described under cellulose^ may act on pentosans seems possible, but there is no direct evidence that such is the case. On the contrary. Cross and Bevan (105) believe that pentosans once formed in the plant, remain thenceforth unaltered. ToUens and Glaubitz (159) assert that the pentosans do not undergo lactic or butyric acid fermentation, and are otherwise unaffected by yeast, as has also been shown by Lintner and Diill (125). The pento- sans are very resistant toward the action of bacteria. Slowtzoff (154) found that a small amount of pure xylan in a putrefying mixture. ^Cf. Czapek, Biochemie der Pflanzen. -Cf. Biedermann and Moritz (34), Brown (35), Brown and Morris (36), Berg- mann (32), Griiss (184), Newcombe (64). 276 Mary Davies Swariz, kept at a temperature of 40 "^ C, did not entirely disappear from the solution before the ninth or tenth day. Two widely distributed fer- menting agents acting on hemicellulose (Bacillus aster osporus Arth. Meyer, and Bacillus clostridieforme, Burri and Anker smit), studied b}' Ankersmit (89), are said by him to occur in insufficient numbers to make their activity of any significance in the alimentary canal of the cow. PENTOSANASES IN LOWER ANIMALS. Extensive investigations regarding the occurrence of pentosan- splitting enzymes in lower animals, have been made by Selliere since 1905. The secretion of the hepato-pancreas of the common snail {Helix pomatia) not only digests cellulose in vitro, '^ but also xylan, ac- cording to this writer (148). In feeding experiments, analyses of the food (oak wood) and excreta of these xylophages showed a higher per- centage of xylan in the former than in the latter (149). Hence xylan must have been digested. In 1907, he showed that pentoses were actually Uberated and absorbed, by testing the blood of these snails, which gave the phloroglucin reaction (151). That sugar can be found in their blood is denied by Couvreur and Bellion (99), but this Selliere attributes to the fact that the sugar content is much less than in higher animals, and hence has been entirely overlooked. Xylanase also occurs in other species of snail (150) such as Helix aspera Miill., Helix nemoralis L., Limax arborum Bouck., Limax variegatus Drap., Arion rufus L., Patella vulgata L., Littorina lit'orea L., Littorina littoralis L., and in a representative of the Coleoptera, Phy- tnatodes variabilis L. The presence of a xylanase in Patella vulgata and the Littorinae is especially significant, as their food consists in pentosan-rich algae. Selliere (150) and Pacault (130) have independ- ently discovered a xylanase in the salivary glands of Helix pomatia. According to Rohmann(134), Aplysia, which subsist largely upon Ulva lactuca, do not, digest the soluble methyl-pentosan (rhamnosan) present in this alga. He finds this carbohydrate present in the glands of the midgut, but regards it as a food residue. PENTOSANASES IN HIGHER ANIMALS. There have been only a few investigations as to the presence in higher animals of enzymes hydrolyzing pentosans. Slowtzoff (154) ^Cf. Biedermann and Moritz (34). Nutrition Investigations. 277 found that pure xylan was not digested by saliva, gastric or pancre- atic juice, but could be gradually hydrolyzed (in two or three days) by 0.2 per cent hydrochloric acid. Bergmann (91) digested pure x>4an with extracts of the intestines of many animals (hen, goose, guinea- pig, sheep, ox, horse), and of the vermiform appendix of rabbits, but in no case found a xylanase. These experiments were performed with suitable antiseptics and controls in all cases. An old experiment by Fudakowski (112), attributing an inverting action upon gum arabic to pepsin, and another by Schmulewitsch (144), attributing such an action upon crude fiber to pancreatui,must be disregarded, as no anti- septics whatever seem to have been used. According to SeUiere (152), neither the pancreatic juice of rabbits, nor a mixture pancreatic and in- testinal juices, will hydrolyze xylan. Negative results were also ob- tained by him with macerated intestines of these animals. On the other hand, chloroform extracts of the intestinal contents of rabbits and guinea-pigs fed fresh hay and bread, produced pentoses in a 5 per cent xylan solution after 48 hours digestion at 37 degrees C, while negative results were obtained with boiled controls. This indicates that the enzymes causing hydrolysis were of bacterial origin, a conclusion sub- stantiated by later work of the same author (153). No xylanase was detected in the excreta of carnivora such as the lion, panther, and wolf. From a centrifugalized extract of human faeces and soluble xylan, di- gested imder aseptic conditions, xylose was obtained after 15-20 hours ; but in meconium of calves and human beings in which bacteria were absent no xylanase could be found, although the intestinal glands were functioning. McCoUum and Brannon (126) have shown that in the case of the cow intestinal bacteria destroy pentosans under anaerobic conditions, the degree of destruction var3dng with the kind of plant. Corn, wheat and oat feeds were incubated with fecal bacteria of this animal, and digestions continued 14 days in atmospheres both of car- bon dioxide and hydrogen, with the following average results : MATERIAI.. ATMOSPHERE. PER CENT OF PENTOSANS DIS-iPPEARING. Corn Fodder CO2 H CO2 H CO2 H 51.78 Corn Fodder Wheat Straw 76.13 28.09 Wheat Straw 37.99 Oat Straw 30.66 Oat Straw 54.00 From this review it is evident that the presence of pentosanases in the higher animals has not yet been demonstrated. 278 Mary Danes Swartz, DIGESTION AND UTILIZATION OF PENTOSANS BY ANIMALS. In the case of men and animals subsisting on a mixed diet, the hex- oses and their derivatives so overbalance the pentosans, under normal conditions, that the utilization of the latter is a question of theo- retical rather than of practical importance. But in the case of herbi- vora, limited to a diet in which pentosans occur in considerable amoimts, the extent of pentosan utilization becomes a question of economic importance. It is not surprising to find, therefore, that since the development of satisfactory methods of quantitative deter- mination, a considerable number of investigations have been made upon such utilization by animals. The results of these experiments are shown in tables on pages 274 and 275. The results in these experiments were obtained by analysis of food and faeces. Lmdsey (123) Gotze and Pfeiffer (113) and ToUens (157) found no measurable amount of pentoses or pentosans excreted in the urine of sheep, but Neuberg and Wohlgemuth (128) state that pento- sans always occur in the urine of rabbits, only disappearing when the vegetable diet is compensated by pentose-free material. They report that 9 per cent of soluble araban (cherry gum) fed to rabbits was ex- creted in the urine. Slowtzoff (154) found 1.4-4.5 per cent of xylan in the urine of rabbits, but no reducing sugar. He also found that if the animal were killed shortly after xylan feeding, xylan could be de- tected in blood, liver and muscles. Hence xylan must have been ab- sorbed from the digestive tract. The feeding experiments show that herbivora digest, on the aver- age, 55-60 per cent of the pentosans in their diet, but since no animal enzymes hydrolyzing pentosans have been demonstrated, and there is- always the possibility of bacterial decomposition in the intestines, the most conclusive experiments as to the actual nutritive value are those of Kellner (118) with the respiration calorimeter. From the slight difference in loss of potential energy, when the furfurol-yielding rye straw preparation was substituted for starch, he concludes that furfurol-yielding substances participate in the formation of fat in the animal body. DIGESTION AND UTILIZATION OF PENTOSANS BY MAN. We have seen that pentosans can be digested by herbivora to a considerable extent. Can they be digested by man? The only feeding experiments on record are by Konig and Reinhardt (120). Nutrition Investigations. 279 In 1902, they conducted researches on two men whose main diet con- sisted of meat and butter or other fat, and beer; to this, in the various experiments, were added respectively (along with sugar, butter, beef extract, etc., used in preparing them) the following substances: Experiment I. Green Peas. Experiment II. Ripe Shelled Peas. Experiment III. Red Cabbage. Experiment IV. Canned White Beans. Experiment V. Soldiers' Bread. Experiment VI. Graham Bread. From analyses of food and faeces the following results were obtained; TOTAL PENTOSANS IN GRAMS. EXP. I. n. m. IV. V. VI. In Food 15.55 0.79 5.08 7.47 23.15 0.59 2.55 3.24 14.01 0.70 5.0 7.75 12.80 1.12 8.75 14.32 52.64 8.66 16.45 20.24 41 26 In Faeces 4.06 9.84 12 97 Percent not utilized, estimating Pen- tosans in Beer as unutilized Total per cent not digested Hence we see that of the total pentosans in the diet 3.24-20.24 per cent were excreted. Only a little furfurol-yielding substance was found in the urine. From the small percentage recovered in these experiments, Konig and Reinhardt (120) conclude that the pentosans are to a high degree utilized by man, but they take no account of pos- sible destruction by bacteria. ^ Since pentosans do disappear from the alimentary tract of men and animals, it behooves us to consider whether, on the assumption that they are hydrolyzed like starch, the pentose sugars so produced are as well utilized as dextrose. Konig and Reinhardt (120) found some furfurol-yielding substance in the urine, and Blumenthal (93) observes that after eating huckleberries, cherries and prunes, pentosans are excreted, but no reducing sugar. Cominotti (100) finds pentoses ab- sent from the urine of man on a meat diet, but always present on a mixed diet. He agrees with Konig and Reinhardt that the output in the urine is small compared with the amount of pentosans in the food, and proposes to investigate the possibility of glycogen formation from pentosans. The behavior of pentoses in the body has been exhaustively reviewed by Neuberg(127).2 It appears from the work of Cremer (102, 103), ^ Cramer (101) has shown (according to a recent review, the original paper was not accessible) that bacteria are essential to hemicellulose transformation. ^For a recent discussion of the absorption and utilization of pentoses see A. Mag- nus-Levy, Oppenheimer's Handbuch der Biochemie der Menschen und der Tiere, 1909, Vol. IV, pp. 395^07. 280 Mary Davies Swartz, 00 O (M C3 »0 (N cj s "S T— 1 d 00 -* iC> Oi i-l 05 O Loss of pote 14.0% Loss of pote CO »n t^ lO CO •<*< "3 .yy o O ^ W h^ 1-:! ^ a) 1/3 >^ Tn «^ ^ ^ ^ bD CO ■^ .S "^ 2 -S ^ .a dj > H fe <1 rG >. M -Sao PP ^ 0) O) (D >» Oi Gi &i 3 c3 rt rt rt O t^ Cm M p^ psi fv^ O W s i o g •S _ I) 4) Qi JS D< M O W OJ O) (jh C/2 (M CO lO Cl 05 C2 CO CO CO o> o 00 00 lO '^ >— 1 l-l '^— ^ rt « 0) OJ -iaj c rH O fc! t3 P-i .5 =o (73 IJ^ w 1 « > en (72 t^ Nutrition Investigations. 281 C5 I— I S 2 a^ (M CO 05 O rj g +J . . . -4^0 S^ (M 05 00 lO -K- -^ ii "M (M CO Tfi ^ .ti 11 Ml _2 oi — < tn d ^ t> 4- cS 1-. T^ tj +-> _j tn ^ m ^ OJ rt S g Pi ^ H "^ < pq U Q 1 °" Id tn ui ^ g 1 o 0) ^ fTi ffl T3 03 o >> i-i r1 ^ rt CJ ,^ l-l a tl) > o ^ m bU c3 OJ o aJ i^ f^ ^ c^ Ji a a a " u u u cj O O O O :S>^^op^ouu CJ Pi 5 I N ^ ^ u M 282 Mary Davies Swartz, Ebstein (109), Frantze (111), Neuberg and Wohlgemuth (128), Sal- kowski (137), V. Jacksch (117),Lmdemann and May (122),Brasch (96) and others, that the pentoses and methyl-pentoses (rhamnose) are ex- creted more readily than the hexoses; that they exert an unfavorable effect in diabetes; and that there is no evidence of their acting as gly- cogen-formers in man. Consequently, even if further experiments justify Konig and Reinhardt's conclusions, the pentosans must appar- ently still play a very small part in the nutrition of man. Occurrence and Nature of Galactans. Next to the pentosans, no hemicelluloses seem to be so widely dis- tributed as the galactans; both occur together in the plant cell, and often in a more or less intimate chemical combination. The pure galactans, i.e., those yielding exclusively galactose upon hydrolysis, have been differentiated into several classes, chiefly by difl'erences in siylubihty or specific rotation, namely: 1. cK-galactan, so named by Mtintz (199), the first to identify galactan as an anhydride of galactose; it composes 42 per cent of luzerne seeds and occurs also in beans, barley, and malt. 2. /3-galactan, isolated from the Hme residues in the sugar beet industry by Lippmann (192). 3. 7-galactan, first isolated from Chinese moss {Sphaerococcus lichenoides) by Payen (262), in 1859, and by him called "gelose." He also identified it in agar-agar^ {Gelidimn corneum) and other algae. The carbohydrates of agar-agar were again studied by Reichardt in 1876, who obtained a substance of the formula Ci2H220n and con- sidered it identical with the "pararabin" which he foiuid in carrots and beets.= In 1881 and 1882, Greenish (180, 181) investigated the carbohydrates of Fucus amylaceus (Ceylon agar-agar) and obtained on hydrolysis a sugar-yielding mucic acid (galactose). From Sphaero- coccus lichenoides he also obtained a substance resembling Payen's "gelose." In 1884, Bauer (169) showed that agar-agar yields galac- tose; and in 1905, Konig and Bettels (190) gave the following per- centage composition of Japanese agar-agar from Gelidium: Per cent. Per cent- Galactans 33 Ash 3.5 Water 20 Pentosans 3.1 Protein 2.6 Crude fiber _..._. 0.4 ^The term agar-agar is applied to the hot water extract of various red algae, mainly species of Gelidium. ^See Pentosans, p. 268. Nutrition Investigations. 283 Another species of marine algae in which galactan has been fully identified, is Chondrus crispus (Irish moss). This is also a red alga. C. Schmidt (210) first examined it, in 1844; he demonstrated that the gelatinizing substance was a carbohydrate and yielded sugar on hydrolysis. Fliickmger and Mayer (178), m 1868, discovered that the water extract of this alga yielded considerable mucic acid. In 1875, Bente (171) obtained levulinic acid from the products of its hydrolysis, and in 1876, reported that it jdelded a non-crystallizing syrup (172). The first quantitative analysis was made by Hadike, Bauer and Tollens (185), who showed that the water extract yielded mucic acid corresponding to about 25 per cent of galactan. Sebor (220), in 1900, foimd in the products of hydrolysis, glucose, fructose and a small quantity of pentose. These observations were verified by Miither (200) in 1903, who further identified the galactose as a (^-galactose. From the large yield of mucic acid, the water extract of Chondrus may therefore be regarded as chiefly galactan, together with some dextran and levulan, and a very little pentosan; groups which, according to Hadike, Bauer and Tollens (185), may be partly or entirely bound into ester-like compounds. Examples of galactans occurring in combination, or close associa- tion with other hemicelluloses are numerous. Lupeose, from luzerne seeds, originally called |3-galactan, yields 50 per cent galactose and 50 per cent fructose (214). The tuberous roots of Stachys tuberifera contain a soluble crystallizable carbohydrate yielding 37 per cent mucic acid, along with an miidentified sugar (225). Para-galactan {galacto-araban) forms a large proportion of the reserve material of many seeds.i Rothenf usser (204) finds that the mucilaginous extract of flaxseed yields equal parts of pentosans and hexosans, the latter being mainly galactose. Galactans and pentosans, as already indicated,^ occur together in many lichens and algae, and also in the pectins.^ Herissey (187) has shown that the "galactine" of Miintz (199) yields equal parts of galactose and mannose. Galacto-mannans also fre- quently occur in the reserve material of seeds, as in those of the date and other species of palm, and in coffee beans; in the American honey iCf. Schiilze (215), Schulze, Steiger and Maxwell (217), Schulze and Castoro (218), Castoro (176), and Goret (179). Also Schulze and Godet, Zeitschrift fiir physiologische Chemie, V. 61, p. 279, for a very complete review of the work of Schulze and his pupils. 2See Chemical Nature of Lichens and Algae:- Konig and Bettels (8), Escombe (6), K. MuUer (11), Ulander (26). ^Cf. Pentosans, p. 268. 284 Mary Davies Swarlz, locust {Gleditschia triacanthus), Goret (179) found the albumen to jdeld 66-70 per cent galactose and 22-23 per cent mannose; he has shown, in fact, that the carbohydrate reserve of almost all seeds with horny albumen consists largely of a mixture of mannans and galac- tans.i GALACTANASES IN THE VEGETABLE KINGDOM. The hydrolysis of the paragalactan of lupine seeds during germina- tion was first observed by Schulze and his co-workers. That ordi- nary diastatic enzymes do not form sugar from the para-galactan of Lupinus hirsukis was demonstrated by Schiilze and Castoro (218). Ptyalin, pancreatin, malt diastase and"taka" diastase, will, however, in the course of 5 or 6 days' digestion at 35^0° C. render this carbo- hydrate soluble in water to the following extent; Per cent. Per cent. Malt diastase 38 Ptyalin 40 Taka diastase 35 Pancreatin 15 Griiss (184) has made exhaustive microchemical investigations upon the germinating date endosperm, in which he has been able to observe the solution of the galactans by enzymes developed during germina- tion. Bourquelot and Herissey (174) find a soluble enzyme hydrolyz- ing galactan,^ produced by the germinating embryos of the seeds of the carob, Nux vomica, fenugrec and luzerne. Shellenberg (208), studying the action of moulds on hemicelliiloses, foimd at least four different ferments showing considerable specificity in their action; seeds of Lupinus hirsutus (containing paragalactan) were attacked by most of these moulds {Mucor neglectus, Mucor piriforme, Rhizopus nigricans, Thamnidium elegans, Penicillium glaucum). Similarly, Herissey (187) found galactose produced from manno-galactans by Aspergillus niger and Aspergillus fuscus; Saiki (205) obtained sugar from Irish moss by digesting it with inulase prepared from Aspergillus niger and Penicillium glaucum; and with "taka" diastase prepared from another mould, Eurolium oryzae. Little is known of the action of bacteria upon galactans. Gran (182) found sugar produced from agar-agar by Bacillus gelaticus, through the action of an enzyme which he calls " gelase." Saiki (105), ^Cf. Mannans, p. 283; for a further discussion of the occurrence of Galactans see V. Lippmann, Chemie der Zuckerarten, Vol. I, pp. 686-697. 2Cf. Mannans, p. 284. Nutrition Investigations. 285 in experiments with B. coli communis, on culture media containing different kinds of comminuted seaweed, found a slight gas production in one culture, in media with agar-agar and Irish moss. GALACTANASES IN THE ANIMAL KINGDOM. The only discovered instance of a galactanase in lower animals is cited by Bierry and Giaja (173), who found that the hepato-pan- creatic juice of Helix pomatia produced galactose from extracts of carob seeds {Ceratonia siliqiia) ; later experiments upon agar-agar. with extracts from a number of crustaceans {Astacus fiuviatilis Rondel, Homarus vulgaris Bel., Maja squinado Rondel., Carcinus moenas L., and Platycarcinus pagarus L.) were entirely negative; the galactans of luzerne and f enugrec were attacked with difficulty by the extract from Astacus. Strauss (221) could find no enzyme attack- ing agar-agar, in the larvae and puppae of various species of Lepidop- tera and Diptera. No galactanases have been found in higher animals. Bierry and Giaja (173), using extracts of luzerne seeds, got negative results with digestive juices of dogs and rabbits, andSawamura (207) ob- tained similar results with extracts of different sections of the alimen- tary canal of swine and horses. Saiki (205) found saHva, pancreatic, and intestinal juices unable to hydrolyze Irish moss, DIGESTION AND UTILIZATION OF GALACTAN BY ANIMALS AND MAN. The first study of the digestibility of galactans in higher animals was made in 1903, by Lindsey (191). Alsike clover-seed, containing 8 per cent galactan, was fed in connection with hay, the digestibility of which had been previously determined; from analyses of food and faeces, the galactan in the hay (1.72 per cent) was found to be 75 per cent digestible, and that in the clover 95.78 per cent digestible. Saiki (205) fed agar-agar and Irish moss to dogs and recovered a large part in the faeces, as shown by the increased amoimt of carbohydrate excreted. Lohrisch (194) fed dogs and rabbits agar-agar in its usual form, and also " soluble-agar " prepared from ordinary agar by Dr. Karl Dieterich of Dresden, Director of the Helfenberg Chemical Fac- tory. This product seems to be partially hydrolyzed in its prepara- tion, since it is not only readily soluble in water, but has slight reduc- ing action; it yields on boiling with Fehling's solution, 3.5-4:.l per cent sugar, and ff a watery solution is allowed to stand 18 hours at 286 Mary Davies Swartz, 37° C, it is further hydrolyzed and yields then 16.9-20.4 per cent sugar. The results of Lohrisch's experiments appear in the folowing table: ANIMAL. FOOD. HEIUCELLULOSE EQUIVALENT OF AGAR FED. HEinCELL- UtOSE EXCRETED. HEMICELL- UI.OSE DIGESTED. Rabbit: Rabbit II Ordinary agar Ordinary agar Soluble agar (given in 9 days) Same as III 18.77 = 14.48 11.8 = 9.11 95.9 = 65.02 53.0 = 35.9 7.1 4.71 14.2 11.7 Per cent. 50.9 48.3 Rabbit III Dog 78.1 67.3 Lohrisch (194) has also studied the utilization of agar-agar in starv- ing herbiVora. In two experiments, rabbits starved for two days were fed ordinary agar as long as they would eat it, other animals of the same weight being kept in starvation as controls ; in a third expe- riment, " soluble agar" was fed. Urine and faeces were collected and analyzed. Of the ordinary agar, about 50 per cent was excreted in the faeces; of " soluble agar," about 25 per cent. No positive evidence of any change in nitrogen excretion attributable to the agar fed, can be drawn from the protocols. One animal died through accident, another survived its control but one day, and the third, in spite of its apparently good digestion of the "soluble agar," died a week before its control. In the case of rabbits made diabetic with phlorhizin and then fed 20^0 grams of both ordinary and soluble agar, Lohrisch (194) found that the D : N ratio remained fairly constant throughout each experi- ment, showing no marked increase in sugar excretion. We see, there- fore, no grounds for assuming that agar-agar {galactan) forms glycogen in rabbits. The first studies on the utilization of galactan by man were made by Saiki (205) (1906). In feeding experiments in which various car- bohydrates were at different times added to a uniform diet, consisting of 513 grams beefsteak, 500-600 grams bread, 40 grams sugar, 31 grams butter, 2 eggs and 2 apples — a diet on which over 98 per cent of the carbohydrates were digested, he obtained the following results: Nutrition Investigations. 287 STJBSTANCE ADDED TO DIET. EQUIVALENT OS SUBSTANCE IN DEXTROSE. CARBOHYDRATES IN FAECES CAL- CULATED AS DEXTROSE. HEMICELLULOSE DIGESTED. 20 grams agar . . . 24 grams agar . . . 40 grams wakame 45 grams kom.bu. Crams. 10 12 4.7 11.4 Grams. 9.2 8.8 3.4 2.5 Per cent. 8 27 28 78 Lohrisch has also studied the digestibility of "soluble agar" in man. Sometimes it is not well borne, especially if given in quanti- ties over 50-60 grams per day and causes gas formation, diarrhoea, and other intestinal disturbances; in other cases, large amounts (100 grams per day) cause no unpleasant symptoms whatever. The agar was dissolved in some beverage, and the diet was otherwise carbohy- drate-free. Some of the results are shown in the following table (194) : DURATION OE EXPERIMENT. AMOUNT DIGESTED. HEMICELL- ULOSE DIGESTED. HEMICELL- NO. As Air Dry Soluble Agar. As Hemicel- lulose. HEMiCKJLLULOSE EXCRETED. ULOSE DIGESTED. Grams. Grams. GrOOTS. Grama. Per cent. 1 1 day 100 61.9 46.06 15.84 25.6 2 1 day 100 61.9 39.1 22.8 36.8 3 3 days 235 145.4 90.5 54.9 37.7 4 3 days 240 148.5 40.8 107.7 72.5 5 1 day 100 61.9 25.4 36.5 58.9 6 1 day 110 67.8 23.4 44.4 65.5 No. 4 was a case of chronic constipation; the high percentage of hemi- cellulose digested is in accordance with the observations of Lohrisch (193) and Pletnew (203), on the extraordinarily good utilization of all foodstuffs in chronic constipation. Two of these experiments were on diabetics, and showed that the 18.36 grams of "soluble agar" ab- sorbed per day caused no increase of sugar in the urine, and had no noticeable effect on nitrogen metaboUsm. From these experiments, we see that ordinary agar is digestible to a very small extent, and that even when changed to an easily hydro- lyzed form, it is only digested to about 50 per cent. Is the part digested absorbed and utilized as galactose? The recent exhaustive 288 Mary Davies Swartz, discussion of the behavior of galactose in the animal body by Brasch (175) renders any details on the utilization of this sugar unnecessary. Hofmeister (188) showed that of all sugars it is most readily excreted. That galactose can form glycogen in dogs and rabbits, has been shown by Weinland (226), Kausch and Socin (189), Cremer (177), Voit (223), Brasch (175), and others.^ Brasch (175) has shown that the assimila- tion limits for galactose lie, for normal man, between 30 and 40 grams, while for dextrose they lie between 100 and 150 grams. Voit (224), Sandmeyer (206), Bauer (170), and others have shown that galactose, even in small amounts increases the sugar excretion in diabetes. It would seem, therefore, that if soluble agar were absorbed as sugar, it would increase the sugar output in the urine. To throw some light on this problem Lohrisch (194) has conducted three respiration ex- periments on men after ingestion of 100-110 grams of soluble agar, of which, on the average, about 63 per cent was absorbed. The changes in the respiratory quotient are shown in the following table: Respiratory Quotient. NUMBER OF HOURS AFTER INGESTION OF SOLUBLE AGAR. NO. IN FASTING. 1 i 2 3 4 5 6 7 I II III 0.768 0.786 0.739 0.768 0.766 0.835 0.794 0.815 0.860 0.825 0.800 0.770 0.767 0.774 0.735 0.724 0.714 IN FASTING. NUMBER OF HO URS AFTER INGESTION OF SOLUBLE AGAR. NO. 8 9 10 11 12 13 I II III 0.768 0.786 0.739 0.693 0.730 0.703 0.618 0.669 The distinct rise in the respiratory quotient in the fourth hour (beginning in the third hour in Experiment I) would indicate that car- bohydrate was being oxidized, which in this case must come from the agar. The low value in the later hours seems due to the oxidation of fatty acids ;2 that such acids may be formed from soluble agar by bacteria, appears probable also from the intestinal fermentation pro- ^Cf . Magnus-Levy, Verwerthbarkeit der Galactose in normalen Organismus : Op- penheimer's Handbuch der Biochemie der Menschen und der Tiere, Vol. IV, p. 379. ^Cf. respiration experiments described under Cellulose. Nutrition Investigations. 289 duced when large amounts of this preparation are taken. A slight increase in acetone output, shown in the metabolism experiments with diabetics, points to the same conclusion. Perhaps, as Lohrisch suggests, the very slow digestion of the carbohydrate, may enable the organism to utUize the galactose formed, and accoimt for its non-ex- cretion, but this reqiiires further demonstration. According to these experiments by Lohrisch, cellulose and the solu- ble galactan show Httle difference in their physiological behavior. Both can be digested to about 50 per cent. Ordinary agar, as Saiki's experiments show, is largely recovered in the faeces; in fact, a thera- peutic practice which has been recently estabHshed is based upon the recognized indigestibiUty of agar, namely, its employment as a remedy in cases of chronic constipation. It is especially valuable, as Mendel (196) points out, in those cases where the difficulty is due to an ex- tremely complete digestion and absorption of all foodstuffs from the alimentary tract, which causes the formation of dry, hard faecal masses {scyhalla) difficult to evacuate. The agar, remaining imdigested and retaining a high percentage of water, gives bulk and softness to the faeces, and facilitates their daUy elimination. Being resistant towards bacterial action, it causes neither gas formation nor produc- tion of harmful decomposition products. According to A. Schmidt (209), it can be advantageously taken in quantities up to 25 grams per day, part with the breakfast cereal, and part with sauce or cream, at another meal. In view of such facts as these, we are hardly prepared to agree with Lohrisch, that ' Cellulose and Hemicelluloses are readily digested. ' Occurrence and Nature oe Mannans. • As widely diversified in origin and character as the galactans, and very intimately associated with them are the Mannans. They show all possible degrees of solubiHty, from the readily soluble mucilage found in certain legumes, to the completely insoluble " reserve-ceUu- lose," which forms the horny albumen in such seeds as the date, and which was long confused with true cellulose. A few examples will serve to show the diverse places in which man- nans may be found. They occur in yeast :i (258) in algae, as Por- phyra laciniata; (278) in moulds, as Penicillium glaucum; (285) in the leaves and roots of the Japanese plant, Conophallus konjaku (280) ; in the bark and wood of many American trees (272) . iFor further discussion see v.Lippmann, Chemie der Zuckerarten, Vol.1, pp. 641- 649, and Czapek, Biochemie der Pflanzen, pp. 325-329. 290 Mary Davles Swartz, The most extensive study has been given to the mannans of various seeds, in which, as already shown, ^ mannans and galactans seem al- most invariably to occur together. The seeds of the carob tree {Ce~ ratonia siliqud) contain a hemicellulose originally called "caruban" by Efiront (241) (1897), but shown by van Ekenstein (282) to yield mannose, and by Bourquelot and Herissey (232) (1899), (^-galactose. The first elaborate studies of "reserve-cellulose" were made by Reiss (264), who showed that the horny albumen of the seeds of Phytelepas macrocarpa, Phoenix dactylifera and other species of palm, Allium cepa, Asparagus officinalis, Iris pseudacorus, Strychnos nux vomica and Caffea arabica, differed chemically from true cellulose in their color reactions, in the ease with which they can be hydrolyzed, and in yielding, instead of dextrose, a sugar which he called "seminose," but which proved to be identical with Fischer and Hirschberger's (242) previously described mannose. Mannan also occurs richly in the tubers of the many species of Or- chis and Eulophia which are the source of commercial salep. On ex- traction with water, they yield a mucilaginous extract which was first studied by C. Schmidt (270) in 1844, and called by him "salep- bassorin"; on hydrolysis with dilute sulphuric acid he obtained, be- s'des some gummy substance and cellulose, a fermentable sugar which he thought to be dextrose. Mulder (259) considered the salep mucilage a mixture of starch and gum or pectin acids, while Franck (243) thought it a modification of cellulose, and Girand (248) a trans- formation of a starchy substance into a variety of dextrin swelling in water. Pohl (263) by precipitation with neutral salts, distinguished an " a-Schleim " and a "/3-Schleim. " According to Thamm (276) , who has made the most recent investigations, ''a-Schleim" does not occur in German salep. ToUens and Gans (277) showed that on hydrolysis, besides dextrose, mannose or, as they called it, " isomanitose " was formed, but this was shown by Fischer and Hirschberger (242) to be identical with ^-marmose. Thamm (276) and Hilger (254) have shown conclusively, that the starch-free water extract contains an anhydride of mannose only. A very resistant type of mannan occurring in some plants, has been designated as manno-cellulose by Schulze (273). Bertrand (227) finds it taking the place of xylan in the woody tissues of gymnosperms. ^Cf . Schulze and his coworkers, and Goret, under Galactans. Also Schulze and Godet, Zeitschrift fiir physiologische Chemie, V. 61, p. 279, for a very complete Nutrition Investigations. 291 MANNANASES IN THE VEGETABLE KINGDOM. There is very little literature concerning the action of bacteria upon mannans. Sawamura (267) observed that extracts of Hydrangea pa- niculata, used in the manufacture of Japanese paper, which contain mannan (along with galactan and araban) , became liquefied on stand- ing. In bacteriological studies with extracts of this plant, and of roots of Conophallus konjaku, he found that only B. mesentericus vul- gatus dissolved these mannans. The action was greatly facilitated, and sugar formation increased if a certain wild yeast, in itself inactive, were present. Traces of a similar enzyme seem to occur in B. prodi- giosus. In his studies of the action of moulds on hemicelluloses, Schellen- berg (269) found that the seeds of Ruscus aculeata, which yield almost exclusively mannose (237-240), were attacked only by Penicillium glaucum. Herissey (253), using pure cultures and water extracts of cultures of Aspergillus niger (grown on media rich in mannose and ga- lactose to incite the development of mannanase and galactanase), with suitable antiseptics and controls, obtained mannose — and galactose — from seeds of Ceratonia siliqua and Gleditschia triacanthus, and an abundant jdeld of mannose from salep; similar results were obtained with Aspergillus f us cus. As early as 1862, Sachs (266) observed the change of the thickened cell-walls of the date endosperm into sugar during germination. The cytases producing this change in 'reserve-cellulose' were later care- fully investigated by Reiss (264), Brown and Morris (230), Newcombe (261), Griiss (251), and others. Still more recently, Bourquelot and Herissey have made many studies on the specific characteristics of these plant enzymes. An exhaustive review of the literature on man- nans and the action of enzymes upon them has been published by Herissey (253), consequently this subject will only be reviewed very briefly here. Gruss (251) has demonstrated that the solution of the date embryo (Phoenix dactylifera) is due to a ferment, the product of whose activ- ity is galactan and mannose. Effront (241) (in 1897) attributed the solution of the albumen of carob seeds (called by him caruban) to a "caroubinase," but thought that the product of its activity was not identical with the products of hydrolysis; in 1899, however, Bourque- lot and Herissey (233) showed the possibiUty of obtaining mannose by the action of a soluble ferment derived from these seeds, which they called "seminase." Shortly afterwards, a similar enzyme was 292 Mary Davies Swartz, isolated by them from the seeds of Phoenix canariensis. Herissey (253) has been able to show that seeds of such legumes as luzerne, fenugrec, and common genet have, at least at the time of germination, ferments capable of transforming mannans — and galactans — into their corresponding sugars. Experiments in vitro show that they are not limited to action upon the seeds by whose embryos they are pro- duced; but act on the reserve-cellulose of seeds from very distinct groups of plants. However, the luzerne ferment does not digest all mannans and galactans; it will hydrolyze the mannans of the tubers of the Orchis family (and commercial salep prepared from them), but not those of the albumen of palm seeds. Griiss (251) has also shown that the enzyme of the date endosperm hydrolyzes starch, although this does not occur in the date seed, and that malt diastase works on a-mannan (the soluble mannan of date seeds, according to Griiss) which does not occur in the barley endo- sperm. Griiss considers diastatic enzymes a group working not only on starch, but also on hemicelluloses. Herissey thinks that diastase and seminase are foimd together in varying proportions in barley, legumes, carob seeds, etc., and that neither is a simple ferment, but a "superposition de ferments," and defines "seminase" as a "ferment or group of soluble ferments, causing the transformation of the car- bohydrates of horny albumens of the seeds of Leguminosae into as- similable sugars." Gatin (247) has made further researches upon the natiire of seminase, and states that during the germination of certain seeds whose reserve is in the form of mannan, the presence of mannose is exceptional, but dextrose occurs in abundance. This phenomenon he attributes to a " manno-isomerase," which transforms the mannose, as fast as formed by the seminase, into dextrose. Experiments in vitro seem to indicate that this is a soluble ferment. MANNANASES IN THE ANIMAL KINGDOM. There are only a few instances on record of mansases occurring in lower animals. Bierry and Giaja (228, 229) found that the hepato- pancreatic juice of Helix pomatia was capable of producing mannose from extracts of carob seeds and salep; that of Astacus fluviatilis, Ho- marus vulgaris, and Maja squinado, from the ivory nut {Phytelepas ma- crocarpa), the two latter hydrolyzing it at ordinary room temperature. On the other hand, the mannans of fenugrec and luzerne were hydro- lyzed with difi&culty, or not at all, by very pure gastro-intestinal juice. No mannanase was found by Strauss (275) in the larvae and Nutrition Investigations 293 puppae of Lepidoptera and Diptera. Similar negative results have been obtained with the digestive enzymes of higher animals. Kino- shita (257) f oiind that emulsin and invertin did not hydrolyze the man- nans of Conophallus konjaku and Gatin (245, 246) tried the blood of rabbits, chicken serimi, the pancreatic juice of dogs, the macerated intestines and pancreas of chickens and cattle, upon salep and carob seeds with negative results; on the other hand, Sawamiira (268) re- ports a mannanase in the extracts from different sections of the ali- mentary tract of swine and horses. DIGESTION AND UTILIZATION BY ANIMALS AND MAN. There are also very few records in the literature of feeding experi- ments with mannans. In a paper in the Zeitschrift fiir Biologic, Voit (283) in 1874^ described one by Hauber, who fed a medium sized dog 390 grams of dry salep powder in the course of eight days. The faeces of the feeding period were roughly marked ofi, and Hauber reported no imchanged salep present in them, because there was no swelling in water as with the original powder. Calculations based on the yield of sugar from the faeces on hydrolysis showed that at least 50 per cent of the salep was absorbed. This seems to have been a very crude ex- periment, and cannot be considered of convincing value. In 1879, Weiske (284) fed carob-beans {Ceratonia siligua) to sheep, along with meadow hay, and compared the nutritive value of this ra- tion with one in which the carob-beans (210 grams) were replaced by an equivalent weight of starch, sugar and protein (from crushed peas). The coefficients of digestibility and nitrogen balance were so nearly the same on the two rations, that Weiske pronounced "Johannis- brod" (carob beans) an acceptable and digestible feed for sheep. In 1890, Schuster and Liebscher (274) tried feeding the sawdust of ivory nut {Phytelepas macrocarpa) to sheep, having previously found that it had a favorable effect on cattle. Merino sheep gained consider- able fat when fed oat straw and vetch fodder, plus ivory nut sawdust furnishing 50 per cent of the digestible carbohydrates. The ration, exclusive of the ivory nut, did not 3deld enough energy for such a re- siilt to be possible, hence the latter must have been utiHzed. The coefficient of digestibiUty, both for the nitrogen-free extract and crude fiber of this material, was at the same time shown by Niebling (262) to be 82 per cent for sheep. ^This paper reviews the early literature on gums. 294 Mary Davies Swartz, From these experiments, mannan would seem to be well utilized by herbivora. The only experimental data regarding the nutritive value of mannans to man, are cited by Oshima (15) from work by Kano and lishima (255), who found the coefi&cient of digestibility of konjaku 82 per cent (prepared from Cowo/>/?a^?M5 konjaku). Further investigations seem highly desirable, in view of the fact that in certain regions food stuffs like salep and konjaku, consisting of almost pure mannan, are among the chief articles of the poor man's diet. It is also a question whether the nutritive value of bark, especially of coniferous trees, is due to mannan present. According to Dillingham (239) the quantity of mannan present does not justify such an assumption, aside from the question of its digestibility. We have finally to inquire whether mannan can be hydrolyzed within the organism, and if so, whether the mannose produced can be retained and form glycogen. From the literature on the subject, it appears that mannose is well utilized by rabbits, dogs and men. Ac- cording to Neuberg and Mayer (260), the d-iorvo. is better utilized than the I- or i-form. Mannose is readily converted to dextrose in the organism; thus Neuberg and Mayer found that a rabbit, receiv- ing 10 grams of /-mannose per os, excreted 1 gram /-mannose and 4-5 grams /-glucose; 10 grams of c?-mannose given rabbits per os, or sub- cutaneously, were almost completely oxidized. Rabbits fed 30 grams {^-mannose by Cremer (238) excreted 3-4 grams in the urine, and dogs given 20 grams by Rosenfeld (265), excreted over 4 grams. This is somewhat more than would be excreted on giving equally large quan- tities of dextrose or levulose. Cremer (238) found no sugar in the urine of a man after feeding 3-12 grams of mannose. That mannose can act as a glycogen former in rabbits, has been demonstrated by Cremer (238) and also by Rosenfeld (265). Neu- berg and Mayer (260) found only a small amount of glycogen in the livers of starving rabbits after feeding /-mannose, but even this form is utilized to some extent. There is good reason for assuming, there- fore, that if mannans can be converted into mannose in the process of digestion, they may be considered as true nutrients for the organ- ism, the mannose being to a high degree capable of absorption and con- version into glycogen. Nutrition Investigations. 295 OCCUERENCE AND NATURE OE LeVULANS. A number of polysaccharide carbohydrates yielding levulose on inversion have been described. They are all levo-rotatory, more or less soluble in cold water and insoluble in alcohol, and easily hydro- lyzed by dilute acid, but have not been investigated sufficiently to permit any conclusion to be drawn respecting their relation to one another. The most important of these substances and their sources are shown in the following table:* NAME. SOURCE. INVESIIGATOR. Inulin Tubers of dahlia, artichoke, Jerusalem artichoke, elecampane; bulbs of onion, garlic, narcissus, hyacinth, and tube- rose; flowers, seed, etc., of various compositae Tanret (321) Chevastelon (291) Pseudo-inulin Inulenin Helianthin Synanthrin Tubers of dahlia, artichoke, Jerusalem artichoke, elecampane; bulbs of onion, garlic, narcissus, hyacinth, and tube- rose; flowers, seed, etc., of various compositae Tanret (321, 322) Levulin Tubers of Helianthus tuberosus (Jeru- salem artichoke) Reidemeister (314) and others Phlein Rootstalks of Phleimi praetense (Tim- othy) Ekstrand and Jo- hanson (296) Cerosin Unripe grains Tanret (320) Graminin Rootstalks of various grasses, e.g., Trisetmn alpestre Ekstrand and Johan- son (296) Harlay (301) Triticin Dracaena austrahs and rubra, Triti- cum repens (couch grass) Reidemeister (314) Sinistrin Bulbs of Scilla Maritima (Sea onions Schmiedeberg (318) or squills) Reidemeister (314) Levulan Molasses in beet-sugar industry v. Lippmann (309) ' Cf. V. Lippmann, Chemie der Zuckerarten, Vol. I, pp. 795-807. 296 Mary Davies Swartz The best known member of this group is inulin,i closely associated with which are the four levulans described by Tanret; these seem to be intermediate products between inulin and levulose, all having greater solubility than inulin, but less levo-rotatory power. The other carbohydrates mentioned are also more soluble than inulin, but have higher specific rotation. LEVULANASES IN THE VEGETABLE KINGDOM. Comparatively few studies have been made upon the action of enzymes on the levulans, and these have been for the most part lim- ited to invdin. Certain micro-organisms as B. Coli communis (295), Clostridium pastorianum (328), and several Schizomycetes, decom- pose inulin, but without any production of sugar. Yeast, according to Tanret (321) does not ordinarily ferment it, but Lindner (308) asserts that certain forms of top yeast change it readily. Levulin is fermented by yeast, according to Levy (307), and triticin, in the course of four or five days, according to Reidmeister (314); but it seems probable that the first changes are due to gradual hydrolysis on standing in water, or to other organisms. The effect of vegetable enzymes on these carbohydrates, as far as they have been studied, is shown in the following table: NAME OF LEVULAN. INVERTIN OF YEAST. MALT DIASTASE. "tAKA" DIASTASE. INULASE OF ASPERGILLUS. Inulin -(8) + (1) -(2) -(3) -(4) + (5) -(6) -(3) + (7) Levulin Graminin ... + very slowly (4) Triticin Sinistrin (1) Levy (307) (2) Reidemeister (314) (3) Chittenden (292) (4) Harlay (301) (5) Reidemeister (314) (6) Schmiedeberg (318) (7) Dean (293) and others (8) Komanos (303) Discovery of the best known ferment for any levulan is due to Green (300) who, in 1888, extracted such an enzyme from the tubers of the Jerusalem artichoke {Helianthus tuber osus), and named it "in- ^For description and early literature see Kiliani (302) and Dean (294). Nutrition Investigations. 297 ulase." Subsequently, Bourquelot (289) found inulase in Asper- gillus niger and Penicillium glaucum; and Chevastelon (291) showed that this enzyme would hydrolyze the inulin of the monoctyledons. Dean (293) has studied the properties of inulase exhaustively, and shown that in Aspergillus and Penicillium it exists only as an endo- enzyme. Went (327) has found inulase also in Monilia sitopkila and other Amylomyces. LEVXTLANASES IN ANIMALS. The first instance of an inulase in an animal organism has been cited by Strauss (319). ' In 1908, he reported studies on the enzymes of seven species of Lepidoptera and Diptera, during their various stages of development {Euproctis chrysorrhea, Ocneria disparata, Bom- byneustria, Bomhyx mori, Galleria melonella, Hyponomenta, Calliophera vomitoria), but found inulase present only in the eating larvae of Bomhyx mori and Hyponomenta. No inulase was present in the larvae of these species after they had ceased eating, nor in the pupae and imagines. The results of Kobert (304) in 1903, with extracts of May beetles, cross spiders, scorpions, cockroaches, ascarides, pupae of pine spiders, and house flies, were entirely negative; so also have been the experi- ments in vitro with digestive juices of higher animals, as shown by table on following page. DIGESTION AND UTILIZATION BY ANIMALS. Inulin is hydrolyzed by very dilute acid (0.05-0.2 per cent at 40° C, according to Chittenden), so that its more or less complete inversion by the gastric juice is possible, and has led many to believe that in spite of the negative results obtained with amylolytic enzymes shown above, it might be converted into levulose, and as such be read- ily utilized by the animal organism. It has therefore frequently been recommended for the diet of diabetics, who show a special tol- erance for levulose; in fact, simply because inulin did not reappear in the urine as sugar, when fed to diabetics, its utilization has been as- sumed by many, no account being taken of its possible reappearance in the faeces. This reappearance is well demonstrated in an experi- ment of Sandmeyer (317) in which, after feeding 80 grams of inulin to a diabetic dog, over 46 grams were recovered in the faeces. 298 Mary Davies Swartz, AUTHORITY. DATE. SOURCE OF ENZYME. KIND OF LEVUIAN. RESULT. Komanos (303) 1875 Saliva Inulin Pancreatic juice Inulin — Schmiedeberg (318).... 1879 Saliva Sinistrin — Chittenden (292) 1898 Saliva Inulin — Pancreatic juice Inulin — Bierry and Portier (288) 1900 Macerated pancreas and intestines of dog, rabbit and seal Inulin — BierryandPortier(288). 1900 Macerated pancreas and in- testines of dogs, rabbits; fed three months on arti- chokes to induce formation of an inulase* Inulin — - Harlay (301) 1901 Saliva Graminin — Bierry (286) 1905 Pancreatic juice of dog Pancreatic juice of dog + Inulin macerated intestines of dogs and rabbits Inulin — ^^ Bierry (287) 1910 Pancreatic juice of dog from pancreatic fistula after in- jection of secretin Inulin — Same pancreatic juice added to macerated intestines of dog and rabbit, in slightly acid, slightly alkaline and neutral solutions Inulin — Hepato-pancreatic juice of Helix pomatia Inulin Levulose Enzyme prepared from he- pato-pancreatic juice of Helix pomatia Inulin Levulose Weinland (326) 1905 Extract of small intestine of dog Inulin — ' Cf. Richaud, (326). Attempts to induce glycogen formation in rabbits have not justi- fied the hopes of the dieto-therapists in regard to inulin as a food for diabetics. The earlier experiments were either negative or open to criticism on account of faulty technique. The more discriminating work of recent investigators (Miura [313] ; and Mendel and Naka- seko [312]), has shown that Httle glycogen is formed from inulin, even under the most favorable circumstances. A brief survey of the expe- riments in this field is given in the following table: u bo 03 22 _ '^ ''JO .a o 4J 3 r3 o bo 3 h > 03 CD O fl a S3 :3 K p aj o 60 a '■2 s V 3 -3 ■Tj 0 O ■* •<# (M TtH eo t^ 00 (N o CO c, I QO.iOi-i COOlNi— i-*cO00»Oi-H i-(T-iOO(MTfliCt>.C^stallized out only on cooling, were pale yellow, soluble in hot water only with great difficulty, but very soluble in alcohol, acetone, or pyridin. After four or five recrystallizations from alcohol, they melted at 152° C. and this melting point remained constant after ten or twelve recrystallizations. However, there were very minute points at which melting seemed to occur about 140° C. Under the microscope, clusters of long needles were seen, each with a tuft of small fine needles springing from its very tip. Dissolved in glacial acetic acid, and examined in a 100 mm. tube, these osazones showed no rotation of polarized light. A very white sample of the dulse carbohydrate was used to deter- mine its specific rotation. It contained 7.1 per cent moisture and 1.68 per cent ash. Two determinations were made, one on a 0.6 per cent solution and the other on a 1.0 per cent solution for which the polariscope readings in a 200 mm. tube were respectively —0.90° and —1.52°. The specific rotation, calculated from these readings was therefore [aj^ = —75.2° and —76.2°, or corrected for moisture and ash, [a]^ = -82.4° and -83.6°, average, -83°. The rate of hydrolysis and maximum reducing power were deter- mined as follows: 5 grams of the material dissolved in 500 cc. of 2 per cent hydrochloric acid wxre boiled in the usual way. At the end of two hours, and at intervals of one hour thereafter, 50 cc. portions were removed, neutralized and made up to 100 cc, and the amount of reducing sugar present determined as dextrose. The following results were obtained: TIME OF BOILING. SUGAR AS DEXTROSE. Hours. Per cent. 2 87.2 3 87.2 4 89.4 5 89.5 That the results vary greatly with the concentration, is shown by the fact that a 0.3 per cent solution boiled 5 hours yielded 67.1 per cent of sugar as dextrose. Having established the fact that this dulse preparation consists of pentosans, with the properties described, further investigations into the exact chemical nature of the carbohydrates composing it were not considered within the province of this work. Nutrition Investigations. 313 Hawaiian Seaweeds. Beside the dulse preparation, three seaweeds have been included ia this group which yielded little or no soluble carbohydrates, namely, Limu Lipoa {Haliseris pardalis), Limu Eleele {Enteromorpha intesti- nalis) and Limu Pahapaha {Uha lactuca, etc.). Limu Lipoa. Limu Lipoa contained a small amount of non-muci- laginous carbohydrate, soluble in cold water as well as hot. It was precipitated by alcohol, in which it came down as a white fibrous mass. On hydrolysis, it yielded a dextro-rotatory fermenting sugar; a test with phenylhydrazin acetate for mannose was negative, as were tests for pentosans. The total amount of this carbohydrate was so small as to be almost neghgible, as far as feeding experiments were concerned, hence the original washed material was used, after grinding to a powder in a coffee mill. It contained a very high percentage of inorganic matter because the thalU were so encrusted with calcareous substances, that it was impossible to remove them entirely by washing. This preparation gave a strong furfurol test, and a single quantitative test for pentosans gave the following results: The sample, weighing 1 gram, contained 10.5 per cent moisture and 18.5 per cent ash. It yielded 0.161 grams of phloroglucid, which according to Krober's tables ^ is equivalent to 0.147 grams pentosans, or 14.7 per cent of the crude substance. Tests for starch and reducing sugar were negative. Only a minute quantity of mucic acid was obtained; a quantity too small to purify and determine the melting point. The products of hydrolysis showed slight fermentation, which was doubtless due to the mannan of the water-extract. A determination of the reducing power made in the same manner as already described, gave the results: TIME OF BOILING. SUGAR AS DEXTROSE. Hours. Per cent. 11 Very little 3 14.3 4 ■ 14.7 6 12.9 8 12.8 Limu Eleele. Limu Eleele yielded no appreciable amount of water- soluble carbohydrate, even after boiling 3 or 4 hours. The dried 1 Zeitschrift fiir physiologische Chemie, XXXVI, appendix. SUGAR AS DEXTROSE. Per cent. 16.8 16.9 18.1 16.8 314 Mary Davie s Swartz, seaweed was therefore simply finely ground for use in feeding experi- ments. It gave a strong furfurol test, but yielded a mere trace of mucic acid. Tests for starch and reducing sugar were negative. The products of hydrolysis contained no fermenting sugar. From this it was e\ddent that the hemicelluloses were chiefly pentosans. Determination of the reducing power gave the following results: TIME OF BOILING. Hours. 2 3 4 5 Limu Pahapaha. Ulva lactuca is said by Rohmann (134) to con- tain a water-soluble methyl -pentosan, rhamnosan; but if this occurs in Limu Pahapaha, it must be in very small amount, as an extract of 50 grams of the dried seaweed, made by boiling 3 or 4 hours, gave very little residue on evaporation to dryness. For feeding experi- ments, the dry crude substance was simply ground to a powder. Like Limu Eleele, it gave a strong furfurol test, but yielded no mucic acid. Starch was present, but no reducing sugar. Fermentation with yeast was marked in 12 hours, probably due chiefly to the hy- drolysis of the starch. Determination of reducing power gave the following results: TIME OF BOILING. SUGAR AS DEXTROSE. Hours. Per cent. 2 28.8 4 31.8 GALACTAN PREPARATIONS. Irish Moss. The carbohydrates of Irish moss are, as already noted, readily soluble in cold water, after the salt has been removed from the sea- weed. By allowing the moss to stand for 24 hours in cold water (about 10 liters to 250 grams of dry substance), an almost colorless, semi- transparent, mucilaginous extract was obtained. By straining this off through gauze, and allowing it to stand over night, for minute particles of cellulose held in suspension to settle, a solution almost entirely free from insoluble material was obtained by decantation. Nutrition Investigations. 315 This was considered sufficiently pure for feeding experiments, and was quickly dried by pouring into broad shallow dishes and placing over a steam radiator. It formed yellowish, translucent scales, which were easily removed, and finely groimd. Subsequent extractions were made in a steam steriHzer, heating several hours at a time. Tests showed that the carbohydrate was not hydrolyzed by this repeated subjection to high temperature. The several extracts were first strained off through gauze and then filtered hot through cotton, to remove the cellulose particles. As these clogged even cotton filters very rapidly, it was found most satisfactory to let the extracts stand over night, decant off the supernatant fluid as far as possible, and filter in a water- jacketed ftmnel. Solutions containing over 1 per cent dry substance could not be filtered through paper. For experi- ments where a perfectly clear fluid was desired, a | per cent solution was filtered hot through plaited paper, and then concentrated on a water bath to the desired strength. One per cent solutions formed a soft jeUy on cooling; 2 per cent solutions, a firm jelly. Even when evaporated to a thick syrup, the carbohydrates of the Irish moss extract are not readily precipitated by comparatively large volumes of 95 per cent alcohol, but form a voluminous, transparent, gelatinous mass. This was found to be more or less characteristic of all the galactans examined. They could be brought down most satisfactorily by addition of sodium chloride to the extract before pouring it into the alcohol. In this way a white precipitate of fine fibers was obtained from the moss. The carbohydrate could also be precipitated by saturation with potassium acetate, and freed from inorganic salts by dialysis, according to the method described by Pohl (263). It could not be precipitated by Fehling's solution, nor by lead acetate in neutral solution. Owing to the opacity of its solutions, and to the fact that its gelat- inizing property made the use of very dilute solutions necessary, no satisfactory determination of its specific rotation could be obtained. A 0.5 per cent solution, clarified with alumina cream, and examined in a 200 nun. tube, showed a rotation of +0.34°, and other trials gave positive evidence that it was dextro-rotatory. The products of hydro- lysis were also dextro-rotatory, and yielded osazones, which after one recrystalHzation from alcohol, had a melting point of 184°-185° C. The carbohydrate gave a red- violet color with iodine, and con- tained no reducing sugar. A faint furfurol test was obtained. Oxi- dation with nitric acid gave a rich yield of mucic acid. Since Hadike, 316 Mary Davies Swartz, Bauer and ToUens (185), and Miither (200) have already shown that Irish moss contains galactan, levulan, dextran and pentosan groups, these tests were simply verifications of some of their observations. Determination of the reducing power gave the following results: TIME OF BOILING. SUGAR AS DEXTROSE. Hours. Per cent. 2 45.6 3 48.6 4 45.8 Hawaiian Seaweeds. Limu Manauea {Gracilaria cor onopij alia) , LimuHuna (Hypnea nidifica), Limu Akiaki {Ahnfeldtia concinna), Limu Kohu {Asparagopsis sanjordiana), Limu Uaualoli {Gymnogongrus). These five seaweeds all contained soluble carbohydrates, which were extracted by boiling in water in an open vessel over a free flame for two hours or longer, Limu Manauea, Limu Huna, and Limu Akiaki, which consist largely of soluble gelatinizing hemicelluloses, yielded most of these on boiling two or three hours. The extracts were strained off through gauze, filtered hot through cotton, and dried in thin sheets as described for Irish moss. While the preparations were dark colored, and had a decided "sea" flavor, they were not unpleasant, and were used in feeding experiments without further purification. As already stated, the carbohydrates were not easily precipitated with alcohol unless a neutral salt (as sodium chloride) was present. Limu Kohu and Limu Uaualoli contained only a small proportion of soluble hemicelluloses, and this was obtained only after boiling 8 to 24 hours. The extracts were also much less gelatinous in charac- ter. The thalli of Limu Kohu are almost like wire when dry, and remain tough and hard even after many hours' boiling. The extracts of these two species were more readily precipitated by alcohol than the others, but the precipitation was greatly facilitated by adding sodium chloride. The carbohydrate of Limu Kohu was precipitated as a white cheese-Uke cake, floating on the surface, while that of Uaua- loH came down as a mass of coarse white fibers. These precipitates were transferred to absolute alcohol, and after standmg several days, were filtered off, washed with ether and dried at 40°-50° C. The Nutrition Investigations. 317 Kohu preparation should have been dried in vacuo, for it proved to be sHghtly hydroscopic, and instead of remaining a fine white powder, became somewhat brownish. The Uaualoli preparation dried easily to a grayish white, Ught, fibrous mass. Tests for starch and reducing sugar were negative on all these substances. Tests for galactans and pentosans were positive in every case. Three-gram samples of the air-dry preparations of Limu Akiaki, Limu Uaualoli and Limu Kohu respectively yielded 0.53 grams, 0.92 grams and 0.64 grams of mucic acid, recrystallized once from ammonium carbonate.^ The products of hydrolysis in no case contained fermenting sugars. It is evident therefore, that these five preparations from the foregoing Hawaiian seaweeds consisted chiefly of galactans, accompanied by some pentosan-groups. From the frequency with which methyl-pentosans have been shown to occur in all seaweeds previously investigated, it is very likely that they occur in all these varieties and it would be desirable to make tests for methyl-pentosans. Determinations of the reducing power were made, as shown in the following table: SPECIES OF SEAWEED. SUGAR AS DEXTROSE. 1 Hour. 2 Hours. 3 Hours. 4 Hours. Limu Manauea Per cent. 43.6 36.0 Per cent. 41.9 58.4 36.0 Per cent. 44.6 55.6 34.0 Per cent. 39.8 Limu Huna 30.8 Limu Akiaki Slippery Elm. For the preparation of the carbohydrate which forms the mucUaginous extract of shppery elm bark, pieces of the latter were torn into narrow strips and allowed to stand over night in cold water ,2 and then the mucilage expressed by squeezing through gauze. This process was repeated rnitil the bark became a mass of separate fibers. The mucilaginous principle swells in cold water to a transparent jelly, but is soluble only to a very limited extent. It was found impossible to filter it, even through gauze, and therefore, although it contained small particles from the disintegrated bark 1 For method cf . Bull. No. 107, p. 55, Bureau of Chemistry, United States Dept. of Agriculture. 2 It was found impossible to extract the mucilaginous principle in hot water. 318 Mary Davies Swartz, fibers, the carbohydrate was precipitated by pouring the thick slimy mass into about six times its volume of 95 per cent alcohol. After standing some hours, a transparent, gelatinous precipitate settled to the bottom, and was filtered of! through several thicknesses of gauze. Dehydrated by means of absolute alcohol and ether, it formed a gray- ish-brown powder. This was found to be soluble in dilute alkali, and was subsequently purified by dissolving in 1 per cent potassium hy- droxide, filtering through cotton and reprecipitating with 95 per cent alcohol. The product was somewhat Ughter in color than at first, but still far from white. It was soluble in hot Fehling's solution, but precipitable with lead acetate. It gave no color with iodine, although a small amount of starch was present in the original bark. Furfurol tests were faint showing only traces of pentosans, but the yield of mucic acid was large, 0.15 grams of mucic acid being obtained from ] gram of the air dry powder. The products of hydrolysis were dextro-rotatory and contained no fermenting sugars. Hence this preparation consisted chiefly of galactan. A MANNAN PREPAEATION. Since none of the algae which form the basis of these studies yielded mannan, save Limu Lipoa, and that in amounts inadequate for the experiments proposed, this hemicellulose was obtained in soluble form from salep. Both the small, horny dried tubers and the gra3dsh- white powder made from them, were purchased from Schieffelein& Co., New York. A preparation of pure mannan was made in the following way: The tubers were soaked in cold water 24 hours, washed thoroughly and ground in a meat chopper. To this mass, cold water was added in large volume, and the whole allowed to stand over night, then the dissolved mannan filtered ofi through gauze. According to Hilger (254), the extract made in this way should contain no starch. But when the tubers are heated before drying, the starch is made soluble, and in this instance the cold water extract gave a blue color with iodine. 1 Hence subsequent extractions were made with hot water on a water bath, for several hours. The salep swells very much in water so that a very large portion was required to get the mannan all into ^Salep tubers purchased since this work was done yielded only a trace of starch jn the cold water extract. Nutrition Investigations. 319 solution.! Xhe extracts, strained through cheese cloth, were digested 24 hours with malt diastase to free from starch, then concentrated to a thick syrup on a water bath, and poured into three times their volume of 95 per cent alcohol. A voluminous, flocculent, and some- what fibrous, snow-white precipitate formed, which was filtered off, pressed free from alcohol, redissolved in hot water, and reprecipitated, (This was done largely to free it from sugar produced by the diges- tion of the starch.) It was then transferred to absolute alcohol and allowed to stand three or four days, after which it was washed with ether, and dried in a vacuum desiccator. A somewhat coarse white powder resulted, containing 6.94 per cent moisture and 0.74 per cent ash.2 It swelled up very readily in water, but dissolved exceedingly slowly to a colorless, semi-transparent mucilaginous solution, which did not reduce Fehling's solution, and examined in the polariscope, after clarification with alumina cream, appeared optically inactive. However, on reprecipitating the carbohydrate with alcohol, and examining the alcohoUc filtrate, sugar was found to be present in small amount. A solution absolutely sugar-free became optically active. A sample in which the sugar had been removed by fermen- tation with yeast, was used to determine the specific rotation. The following results were obtained: (1) A 2 per cent solution in a 200 mm. tube read —1.59°; applying corrections for moisture and ash, [a]jy = —43.1°. (2) A sample containing in 100 cc. 0.5868 grams mannan dried to constant weight at 105° C. read —0.48°; corrected for 0.4 per cent ash, [al^ = —43.8°. According to Thamm (276), salep extract is inactive. In the above experiments, the levo-rotatory nature of the mannan was at first obscured by the presence of traces of reducing sugar formed by the hydrolysis of the starch, which could not be detected by testing directly by Fehling's solution. Thamm, however, in several ways carefully tested salep hydrolysis products for dextrose with negative results, so that the only way to account for these confhcting results seems to be to attribute it to difference in the specimens of Orchis which furnished the mannan. Salep-extract is readily precipitated by Fehhng's solution in floccu- lent white masses. It is not precipitated by lead acetate in neutral solution (nor, according to Thamm [276], in solutions of other neutral salts), but is precipitated by basic lead acetate. A furfurol test was faintly positive, verifying the report of traces of pentosans by Tollens and Widtsoe (163), and also by Thamm (276). 115 liters of water to 100 grams salep powder, according to Thamm (276). ^Thamm found 0.483 per cent. 320 Mary Dairies Swartz, The products of hydrolysis were dextro-rotatory and contained sugar fermentable with yeast. A rich yield of mannose-hydrazone was obtained with phenyl-hydrazine acetate, melting on recrystalliza- tion at 188° C. According to Thamm (276), salep extract yields ex- clusively mannose on complete hydrolysis. Hydrolyzed for three hours, the reducing power of this mannan was 91.6 per cent. Determinations of ash, moisture, starch, and mannan were made on the salep obtained in the form of a powder. Starch and mannan were determined as follows: 1 gram of air dry powder was boiled in 250 cc. water, and after cooling to 37.5° C, the starch hydrolyzed with malt diastase, dialyzed sugar-free. The solution was then filtered, con- centrated to small volume, and the mannan precipitated with absolute alcohol. The precipitate was filtered off, dissolved in a little water and reprecipitated, to obtain any sugar retained in the first precipi- tation. The mannan was then dried at 100° C. and weighed. The filtrates were combined, freed from alcohol, hydrolyzed with 2 per cent hydrochloric acid 45 minutes to convert all the maltose to dextrose, and sugar determined by Allihn's method. The results of these analy- ses are shown in the following table: Per cent Per cent Moisture. 0.77 Starch 26.4 Ash 8.9 Mannan 19.5 According to Dragendorfi the composition of Orchis tubers is as follows: Per cent Per cent Starch 27.3 Protein 4.9 Mucilage 48. 1 Cellulose 2.4 Sugar 1.2 Thamm also reports a yield of 40^5 per cent mucilage from the salep powder used in his investigations. Hence the powder used in this the present experiment was for some reason very deficient in mannan. Its reducing power w^as as follows: TIME or BOILING. SUGAR AS DEXTROSE. Hours. Per cent 2 74.2 3 75.8 5 75.8 ^Cited in the National Dispensatory (1884), also by Thamm (276). Nutrition Investigations. 321 A LEVULAN PREPARATION. Commercial Squills, consisting of the dried and broken leaves of the bulbs of Scilla maritima (or Urginea Scilla Stenh.) jdeld, as dis- covered by Schmiedeberg (318), the levulan sinistrin. They were finely ground in a coffee mill, and the sinistrin prepared according to Schmiedeberg's directions. To the dry powder sufficient water was added to make a thin cream, and then a saturated lead acetate solu- tion until further addition produced no precipitate. To the clear, straw-colored filtrate, freed from lead with hydrogen sulphide, was added freshly prepared milk of lime, with constant stirring, until a somewhat creamy consistency was produced. To f aciUtate the for- mation of sinistrin-calcium carbonate, this mixture was concentrated on the water bath for some time (as suggested by Reidemeister) [314]. The precipitate was then sucked dry on a Blichner fimnel, washed thoroughly with cold water (being rubbed up in a mortar for the pur- pose), again sucked dry, rubbed to a cream with water, and treated with carbon dioxide imtil the fluid was no longer alkaline to litmus. After heating to facilitate the complete separation of the calcium carbonate, the sinistrin in solution was filtered off, a little oxahc acid carefully added to remove the last traces of lime, and the solution then decolorized with charcoal, and evaporated to a syrup at a temperature of about 40° C. From this solution the sinistrin was precipitated with 95 per cent alcohol, as a white gummy mass. Transferred to abso- lute alcohol, and allowed to stand 24-36 hours it became very tenacious, but on longer standing, with occasional stirring, it grew brittle, and finally crumbled to a coarse white powder, which was dried in a vacuum desiccator. This material was readily soluble in cold water. (According to Schmiedeberg [318], even solutions of 20-30 per cent are not syrup-like.) It gave no color with iodine, did not reduce Fehling's solution, and was not precipitated by it. This preparation, at first, contained 13 per cent moisture and 0.76 per cent ash. De- termination of the specific rotation then gave the following results: A 2 per cent solution in a 200 mm. tube, read —1.32°; corrected for moisture and ash, [a]D = -38.2°. After longer standing (three months) over sulphuric acid, the moisture content was 4.8 per cent, and determination of specific rotation gave the following results: A 1 per cent solution in a 200 mm. tube, read -0.55°; corrected for moisture and ash, [a]D = -29.1°. Schmiedeberg (318) found the average for [a]D = —41.4°, and Reidemeister (314), [a]D = —34.6°. It is impossible to account for these differences. Reidemeister claims 322 Mary Davies Swartz, that the rotation increases on standing, but in these solutions there was no change in 48 hours, at room temperature. On hydrolysis, sinistrin yields a levo-rotatory, reducing sugar, fer- menting with yeast. Schmiedeberg (318) reports this as a mixture of levulose and an inactive sugar, but Reidemeister (314) declares that it is neither a mixture of levulose and an inactive sugar, nor of levu- lose and dextrose, in spite of the fact that he found for it la]D= —88°, while for levulose, [a]D= —106°, a difference for which he is unable to account. SUMMARY. The composition of the preparations which have been described is best shown in the following table: SOURCE OF MATERIAL. Dulse (Rhodymenia Palmata) Limu Lipoa {Haliseris Par- dalis) Limu Eleele {Enteromorpha intestinalis) Limu Pahapaha {Ulva lac- tuca, etc.) Irish Moss {Chondrus crispus) Limu Manauea {Gracilaria coronopifolia) Limu Huna {Hypnea nidifica] Limu Akiaki {Ahnfeldtia con- cinna) Limu Uaualoli {Gymnogon- grus) Limu Kohu {Asparagopsis sanfordiana) Slippery Elm iUlimis) Salep {Orchis.) Squills {Urginea scilla) [Sinis- trin] NATURE OF CARBOHYDRATES PRESENT. Pentosans. Galactan. Mannan. Levulan. Dextran + + + + Trace + + + + + Trace + + + + + + + + + + (Starch) + The foregoing observations correspond with those of Konig and Bettels (8), in that the marine algae all yield pentosans, and fre- quently galactans. The gelatinizing principle in every case appears to be due to the galactan groups. No specific tests have been applied Nutrition Investigations. 323 for fructose, the polysaccharide of which also appears to be common in algae, but the absence of fermenting sugar in all the algae except Limu Lipoa, indicates that if present, it is in too small amount to be detected in the hydrolysis products of 5-10 grams of crude material. The reducing power has been determined on each substance used in feeding experiments; the results of all determinations are summarized in the following table: SUBSTANCE. SUGAR AS DEXTROSE ASTER BOILING. 1 Hour.'2 Hovirs. 3 Hours. 4 Hours. 5 Hours. 6 Hours. 8 Hours Dulse Limu Lipoa Limu Eleele Limu Pahapaha Irish Moss Limu Manauea Limu Huna Limu Akiaki Salep (Powder) Salep (Pure mannan) [Per cent. Per cent. I 87.2 I I 16.8 : 28.8 . 45.6 I 41.9 43.6 ! 58.4 36.0 36.0 1 74.2 Per cent. Per cent J Per cent. 87.3 89.4 ! 89.5. 14.3 I 14.7 I 16.9 ! 18.1 j 16.8 i 31.8 i 48.6 1 45.8 44.6 ' 39.8 55.6 i 30.6 34.0 75.8 91.6 Per cent. 12.9 Per cent. 12.8 75.8 Bacteriological Investigations. INTRODUCTION. It is an accepted fact that even cellulose, with its high powers of resistance, is to some extent decomposed in the alimentary tract by bacteria. It is therefore reasonable to expect that the less resistant hemicelluloses will also be attacked and decomposed by bacteria. The object of these experiments has been to throw some Hght on the problem as to what organisms are most likely to effect such a decom- position, and whether there is an appreciable production of sugar as a result of bacterial activity. The four classes of hemicelluloses under special investigation have been represented by the following sub- stances : Pentosans Dulse. / Irish Moss. ' \ Limu Manauea. Galactans . Mannans Salep. Levulans Sinistrin. 324 Mary Davies Swartz, Both aerobic and anaerobic cultures have been made, in neutral, faintly alkaline, and faintly acid reaction, with solutions made from the carbohydrates alone, and with the addition of small amounts of such nutrients as beef extract or peptone to facilitate the growth of the organisms. Anaerobic cultures in test tubes have been made by the Wright method ; anaerobic ciiltures in Erlenmeyer flasks, by passing a stream of hydrogen through for half an hour, and then sealing hermetically. The aerobes which have been employed all occur in the human digestive tract. Both aerobic and anaerobic cultures from the faeces of human subjects have also been used, in conjunction with soil bacteria from street sweepings. Tests for the presence of reducing sugar have been made by pre- cipitating the carbohydrates in solution with absolute alcohol, evapor- ating the alcoholic extract to dryness, taking up the residue in 2 or 3 cc. of water, and boiling two minutes with Fehling's solution. Suitable controls have been used in all cases. TRIALS WITH PURE CULTURES OE AEROBES. One per cent solutions of the preparations from dulse, Irish moss and salep, neutral, acid, and alkaline in reaction, and consisting of, (1) pure carbohydrate; (2) carbohydrate plus J per cent beef extract and f per cent sodium chloride; (3) carbohydrate plus 1 per cent peptone and i per cent sodium chloride, have been used as culture media. Five cc. portions of each of these solutions were placed in test-tubes with a pipette, and inoculated with the following organisms: B. Coli communis, B. Pyocyaneus, B. Prodigiosus, B. Proteus vulgaris, B. Pyogenes foetidus. To approximate the conditions in ordinary digestion of these car- bohydrates, they were incubated for three days at a temperature of 37.5° C. At the end of this time, nearly all gave evidence of some bacterial growth. Salep-peptone cultures of B. Pyocyaneus showed a brilliant green; salep solutions containing B. Pyogenes foetidus, and B. Coli in alkaline-beef extract media, had changed from trans- parent colorless solutions to an opaque white jelly insoluble in water. The carbohydrates were then precipitated with alcohol, and after standing several days were compared with controls similarly prepared, to see whether any change could be observed in the nature or amount of carbohydrate. The results were in all cases negative. These pre- cipitates were then transferred to small folded filter papers of uniform Nutrition Investigations. 325 weight, previously prepared. The alcohoHc filtrates were tested for sugar; the precipitates were dried, and their weight compared with that of the control. It was thought that this rather crude method would show whether any considerable amount of the carbohydrate had disappeared. The results were so largely negative that weighings of every precipitate were not made. There seemed to be a slight loss of dulse, in some of the cultures of B. Proteus vulgaris, B. Pyogenes foetidus, and B. Coli communis, but repetition of these experiments allowing the organisms in question to grow two weeks, not only in dulse but also in salep media, did not justify any conclusion that an appreciable amount of carbohydrate had disappeared. All tests for reducing sugar were negative. Four per cent solutions of Irish moss, and two per cent solutions of limu manauea were then prepared, with reactions and additions of nutrient material as described in the first series of experiments. These formed firm jelHes, which were used to study the possibility of lique- faction or gas formation. Stab cultures were made, and grown at a temperature of 25°-30° C. for one to three weeks. No hquefaction or gas formation was observed in any case. TRIALS WITH MIXTURES OF AEROBES. Mixtures of B. Pyocyaneus, B. Prodigiosus, B. Proteus vulgaris, and B. Pyogenes foetidus, were used, also mixtures of faecal and soil bacteria. These were first inoculated into nutrient bouillon, the former from pure cultures, the latter from human faeces and street sweepings, and incubated 24 hours. Five cc. portions of these cultures were then introduced into 50 cc. of neutral solutions of each of the different carbohydrates, in small Erlenmeyer flasks, and these cul- tures allowed to grow for four weeks at 37.5° C. At the end of this time, no marked change had taken place save in the salep culture of B, Pyocyaneus, B. Proteus vulgaris, B. Pyogenes foetidus and B. Prodigiosus. This had changed from a colorless, semi-transparent, shghtly mucilaginous fluid, to a firm, white opaque jelly, insoluble in water, but readily soluble in dilute alkali; a phenomenon already observed with this carbohydrate in cultures of B. Coli communis and B. Pyogenes foetidus. No liquefaction had taken place with Irish moss nor limu manauea. The carbohydrates were then precipitated with alcohol, the alco- holic extracts tested for sugar, and the precipitates hydrolyzed by boiling with 2 per cent hydrochloric acid, neutrahzed, made up to a 326 Mary Davies Swartz, definite volume, and examined in a polariscope. The results of these experiments are shown in the following table. Mixtures of B. Pyo- cyaneus, B. Prodigiosus, B. Proteus vulgaris and B. Pyogenes foetidus are designated A, and mixtures of faecal and soil bacteria, B. BACTERIAL REDUCTION OF ROTATION AFTER HYDROLYSIS. CULTURE ' FEHLING'S SOLUTION. Experiment. Control. Dulse B A B :::::+: + 0.13° + 0.20° +0.27° Not determined. Not determined. +0.17° -0.97° +0.20° Irish Moss Irish Moss + 0.20° Limu Manauea .... Salep Salep Sinistrin A A B A +0.20° —0.97° The action of putrefactive organisms upon the dulse preparation was also studied, according to the method used by Slowtzofi (154) in the case of xylan. One hundred grams of chopped lean beef and 10 grams of sodium carbonate were added to 1 liter of water, and the mixture allowed to stand in a warm place for three days. Two himdred and fifty cc. were then removed for a control, and to the remainder 0.5 gram of dulse was added. This solution gave a strong pentosan reaction; the control was pentosan-free. The two solutions were put in a warm place, and tested daily for pentosans. After five days' digestion, the reaction of the dulse solution was very much fainter than at first, but it did not entirely disappear till the twelfth or thirteenth day. Slowtzofi found that xylan disappeared in nine or ten days, but his solution was kept at a temperature of 40° C, while these mixtures remained at a temperature of from 30° to 35° C, a condition less favorable for rapid decomposition. Solutions of Irish moss were digested with faecal mixtures in the following manner: Human faeces were rubbed to a mud with water. Ten cc. portions of this material were added to flasks containing 50 cc. of a 1 per cent "moss" solution, and allowed to digest in a warm place for 24 hours. A portion of water inoculated in the same way was used as a control. Small portions of these solutions were then evap- orated nearly to dr3aiess, extracted with alcohol, and tested for reduc- ing sugar. The results were wholly negative. That limu manauea is not entirely resistant to the action of putre- fying organisms is shown by the following: A solution was made Nutrition Investigations. 327 up to contain 2 per cent of the air dry extract, 1 per cent peptone, J per cent beef extract and \ per cent sodium chloride. This could be filtered through paper only on a hot, water- jacketed funnel, from which it dropped as a clear, amber-colored jelly. After standing imsterilized over night in a warm room, this was found to be entirely broken up by the formation of gas throughout the whole mass. The reaction, which had been neutral, was now acid to litmus. This material was placed in a flask and allowed to stand for two months, at the end of which time, the greater portion was Uquefied, the former lumps of jelly being reduced to smaU particles distributed throughout the liquefied portion. Alcoholic extracts did not reduce Fehling's solution. A sterile preparation of the plain manauea extract in test tubes was inoculated with some of this material, but without produc- ing the same striking results. There were evidences of growth, but none of liquefaction or gas formation, in the course of two weeks. TRIALS WITH ANAEROBES. The action upon Irish moss of pure ciiltures of the powerful putre- factive organisms B. Putrificus, Bienstock, B. Mahgni oedematis, and B. Anthracis symptomatici, was tried in the following way. A 4 per cent solution of the moss was prepared, which would not become Hquefied at a temperature of 30°-35° C. From this material culture media were prepared, neutral, alkaline, and acid in reaction, using the solution plain, and with the addition of | per cent beef extract and \ per cent salt, or 1 per cent peptone and \ per cent salt. Test tubes were inoculated from fresh, active cultures, and the organisms allowed to grow for one to three weeks, being examined at first daily, and later every three or four days, for liquefaction and gas formation. The results were negative in all cases, save that in the peptone media an occasional small bubble was seen, mth cultures of the bacilli of mahgnant oedema and symptomatic anthrax. However, the same phenomena were observed in peptone-agar tubes used as controls. Mixtures of B. Anthracis symptomatici and B. MaHgni cedematis were tried upon solutions of dulse, Irish moss, salep and sinistrin, in the following way: Small Erlenmeyer flasks containing 50 cc. of 1 per cent solutions of each of these carbohydrates, and 5 cc. of ordi- nary nutrient bouillon, were inoculated with fresh ciiltures of these organisms, rendered anaerobic, and incubated for four weeks at 37.5** C. On inspection, no change was apparent. The carbohydrates were removed, the alcoholic extracts examined for reducing sugar, and 328 Mary Davies Swartz, the carbohydrate residues hydrolyzed and examined in the polari- scope, as in similar trials with aerobes. The results are shown in the following table: NAME OF SUBSTANCE. reduction of eehling's solution. ROTATION AFTER HYDROLYSIS. Experiment. Control. Dulse Irish Moss Salep Sinistrin . . Lost by laccident + 0.24° I +0.20° + 0.13° +0.20° - 0.27° - 0.97° Mixtures of soil and faecal bacteria were also tried, the experiments being carried out just as described for mixtures of the bacilli of symp- tomatic anthrax and malignant oedema. The results are shown in the following table: reduction of feeling's solution. ROTATION AFTER HYDROLYSIS. NAME OF SUBSTANCE. Experiment. Control. Dulse Irish Moss + + 0.13° + 0.20° + 0.03° + 0.20° + 0.20° Salep + 0.20° DISCUSSION AND SUMMARY. It seems reasonable to expect, that if the hemicelluloses used in these trials were readily attacked by micro-organisms, there would have been some evidence of change in three days, if conditions for growth were favorable as regards reaction and temperature; but although the concentration of the solutions was moderate, the reaction varied, and temperature 37.5° C, results were negative, even in the cases where nutrients were added to facilitate bacterial growth. Apparently all of the material was recovered in unaltered condition, save in certain instances where salep underwent an insoluble modification. In trials where the cultures were allowed to grow from one to three weeks, no dilf erence in the results could be detected, by the methods employed. In solid media there was no liquefaction and practically no gas formation, except in the case of the peptone-beef extract preparation of limu manauea, on exposure to the air. Nutrition Investigations. 329 Marked evidences of change were observed in one trial with a putre- factive mixture (on diilse), and in some of the four- week cultures. Irish moss was the most thoroughly investigated and proved the most resistant. In the long experiments (4 weeks) where the other carbohydrates suffered more or less change this one remained appar- ently unaltered. The results of this series are summarized in the following table: Irish Moss. CULTURES USED. reduction of fehxing's solution. ROTATION OF UNALTERED CAR- BOHYDRATE AFTER HYDROLYSIS. Irish Moss. Control. Mixture of Pure Aerobes . Mixture of Faecal and Soil Bacteria (aerobic) Mixture of Bacilli of Malignant Oedema and S3Tnptomatic Anthrax . . Mixture of Faecal and Soil Bacteria (anaerobic) + 0.20° I +0.20° i + 0.27° ' +0.20° + 0.24'= + 0.20° + 0.20° I +0.20° The single experiment with the galactan, limu manauea, imder the same conditions, with the mixture of pure aerobes, gave similar results, but the fact that Uquefaction occurred in the peptone-beef extract culture medium after exposure to the air, shows that general conclusions as to the behavior of galactans cannot be drawn from study of a single representation of the class. We have, however, further proof that the galactans are not easily decomposed by bacteria^, in the fact that aqueous solutions of all the galactans included in the present series, could be left several days in the warm atmosphere of the laboratory without any apparent change taking place; and in the fact that agar-agar, so widely used in bacteriological laboratories on account of its indifference to bacterial action, is a member of the galactan group. It has been suggested^ that extracts of other sea- weeds might prove good substitutes for agar-agar as culture media, if fully investigated. So far, the greatest objection to use of Irish moss in this way is that it tends to Kquefy at body temperature; strong solutions (4 per cent) can, however, be kept fairly firm at a iCf. Reed (18). 330 Mary Davies Swartz, temperature of 30° C. The extract of limu manauea is free from these objections, but extensive experiment is still necessary to demon- strate its powers of resistance. The soluble dulse pentosan is certainly decomposed not only by putrefactive organisms under the most favorable conditions {e.g., in meat mixtures), but by aerobes and anaerobes in solutions where the carbohydrate is the chief source of nutriment. The results of the four weeks' digestions are summarized in the following table: Dulse. CULTURES USED REDUCTION OP fehling's solution. ROTATION OP UNALTERED CARBONTDRATE APTER HYDROLYSIS. Dulse. Control. Mixture of Faecal and Soil Bacteria (aerobic) + 0.13° (Lost by Accident) +0.13° +0.20° Mixture of the Bacilli of Malignant Oedema and Symptomatic Anthrax. . Mixture of faecal and Soil Bacteria (anaerobic) +0.20° +0.20° In the present studies, this pentosan stands second to the galactans in degree of resistance. Sawamura (267) thought that he observed a slight hydrolysis of mannan by B. Prodigiosus, an observation which has not been verified in these experiments. No reducing substance was detected in the three-day cultures nor the four-weeks cultures, in which this organism was present. The opaque jelly, insoluble in water, formed from salep by the action of B. Coli communis, B. Prodigiosus, and mixed cultures containing these organisms, resembles an intermediary product of the acid hydrolysis of salep-mannan described by Thamm (276). He isolated and examined two such products, one forming an opales- cent solution in water, the other insoluble, but passing over into the soluble form by treatment with dilute alkali; both were anhydrides of mannose. It seems reasonable to inquire whether this insoluble material produced by bacterial action may not be regarded as an early stage in the hydrolysis of the carbohydrate under consideration, especially in view of the fact that in all the other four-week trials a very definite reduction of Fehling's solution was noted, corresponding Nutrition Investigations. 331 in strength with the loss of unaltered carbohydrate, as shown in the following summary: Salep. CTXLTUKES USED. REDUCTION OP rETTT.TNG 's SOLUTION. ROTATION OF UNALTERED CARBO- HYDRATE APTER HYDROLYSIS. Mixture of Pure Aerobes ! (Insoluble I jelly) Mixture of Faecal and Soil Bacteria ' (aerobic) + Salep. Not det ermined + 0.17° i +0.20° Control. Mixture of the Bacilli of Malignant Oedema and Symptomatic Anthrax . . Mixture of Faecal and Soil Bacteria (anaerobic) + + + 0.13° + 0.03^ + 0.20° + 0.20° These experiments give some grounds for expecting the hydrolysis of salep in the aUmentary tract, through the action of bacteria. Two experiments with sinistrin gave the followmg results: Sinistrin. reduction of fehling's SOLUTION. ROTATION OF UNALTERED CAR- BOHYDRATE AFTER HYDROLYSIS. CULTURES USED. Sinistrin. Control. + -0.97° -0.27° -0.97° Mixture of Bacilli of MaHgnant Oedema and Symptomatic Anthrax -0.97° Sinistrin is therefore hydrolyzed by the anaerobic putrefactive organisms, but further experiments are necessary to determine how readily this change takes place. Physiological Investigations. In the physiological experiments, attempts have been made to answer the following questions: (1) To what extent are hemiceUuloses digested by animal and vegetable enzymes? (2) Can they be ab- sorbed and utUized without intervention of the alimentary tract? 332 Mary Davies Swartz, (3) Do they reappear in the faeces after administration per os? The various experiments will accordingly be discussed in these three groups: (1) Trials with Enzymes; (2) Parenteral Trials; (3) Feeding Experiments. TRIALS WITH ENZYMES. Approximately 1 per cent solutions of the various hemicelluloses (with the exception of Limu Lipoa, which was finely ground and sus- pended in water), have been digested for 24 hours at 37.5° C. in the presence of toluene, with the following enzymes: (1) Filtered human saliva. (2) Malt diastase, dialyzed sugar-free. (3) "Taka" dias- tase {Eurotium oryzae). (4) Chloroform extract of pig's pancreas. (5) Fresh pancreatic juice of dogs. (6) Chloroform water extract of dog's intestines. (7) Glycerol extract of pig's stomach. Digestions have also been made with 0.2 per cent hydrochloric acid, to determine whether any of the action of the artificial gastric juice might be due to the acid present. The activity of the amylolytic enzymes has always been tested first with starch paste, and that of the gastric extract with fibrin. Boiled controls have been employed in every instance, and all trials have been made in duplicate. Tests for reducing sugar have been conducted in the following manner: At the end of 24 hours the solutions were evaporated to thick syrups on the water bath, to free from toluene and to concen- trate so that the undigested hemicelluloses could be readily precipi- tated by absolute alcohol. The alcoholic extracts were filtered ofi and evaporated to dryness; the residues were taken up in a few drops of water and tested for sugar with FehHng's solution. The results of all digestion trials are shown in the table on opposite page. PARENTERAL INJECTIONS. Methods and Technique. Small dogs were used for all injections, after a confinement in cages long enough to obtain samples of normal urine. The carbohydrates employed in these experiments were preparations of dulse,i Irish moss,^ salep,^ and sinistrin.* They were introduced suhcutaneously , by means iCf. p. 303. 2Cf. p. 308. 3Cf. p. 312. «Cf. p. 315. + I + + + ^o + + + + + + 02 c3 CO + .2 O tn C pj cd " rt ;3 1 S w < t^ fc^ -, :i ^ ^ 7i s :i s s s a s p vii ;j 13 '^13 P ^ w OOOUUOOOOC -72 334 Mary Davies Swariz, of a syringe, or intraperitoneally, by means of a needle and burette with pressure-bulb attached, always under aseptic conditions. After receiving injections, the animals were replaced in cages, and the urine collected under toluene. The excess of toluene was removed, at the time of examination, by means of a separatory funnel, and the urine measured, filtered, and tested for reducing substances with Fehling's solution. Qualitative tests for the carbohydrates were made in the following manner: (l) for dulse and salep, by boiling a few drops of urine with Fehling's solution, from which these hemicelluloses were precipitated in fine white flocks, even if only traces were present; (2) for Irish moss, by the reduction of Fehling's solution after hydrolysis of the urine with dilute hydrochloric acid;i (3) for sinistrin, by the marked increase in the levo-rotation of the urine. Isolation of the carbohydrates was accomplished by freeing the urine from inorganic salts with lead acetate, removing the excess of lead with hydrogen sulphide, and concentrating the salt-free solutions to a small volume. Dulse and Irish moss were then precipitated with absolute alcohol; salep with alcohol or Fehling's solution; sinistrin with milk of lime, being freed from its calcium compound by the method used in its preparation. ^ These substances were identified as carbohydrates, by their yield- ing reducing sugar on hydrolysis; salep and sinistrin were further identified by their levo-rotation, Irish moss by testing for mucic acid, and dulse by testing for furfurol. Quantitative determinations of dulse, salep and sinistrin were made by polariscopic examination in a 200 mm. tube, all samples of urine being clarified with equal volumes of alumina cream. A satisfactory quantitative method for the determination of Irish moss was not developed. It proved impossible to estimate any of these carbohy- drates quantitatively by the method of acid hydrolysis. In some instances, especially with Irish moss, a trace of reduction was ob- tained, but in most cases, the results were negative, although the hemi- cellulose was known to be present.^ ^Trial was made of Bauer's method (Zeitschrift fiir physiologische Chemie, 51, p. 158, 1907) of determining galactose in urine as mucic acid, by concentrating 100 cc. of urine with 25-35 cc. of concentrated nitric acid (sp. gr. 1.4) to a volume of 20 cc, but owing probably to the low percentage of galactose from the small amount of Irish moss present, this test was unsatisfactory. 2Cf. p. 315. ^Samples were removed and tested every half hoiir for 2^ hours. At the end of 1 hour they were usually neutral, or slightly alkaline in reaction. Addition of suf- Nutrition Investigations. 335 INJECTIONS OF DUXSE. 1. Subcutaneous. A dog weighing 11 kg. received 60 cc. of a dulse solution contain- ing 0.9 grams of pure substance. No reduction of Fehling's solution was observed at any time. The time and rate of dulse excretion are shown in the following table : Examination of Urine. ESTIMATED EXCRETION OF DULSE.* February 1, 12:30 P.M February 1, 1 P.M Injection February 2, 10 A.M ''■ 226 February 3, 10 A.M 250 February 4, 10 A.M I 150 February 5, 10 A.M ' 210 February 6, 10 A.M : 310 February 7, 10 A.M j -0.14°t - 0.62° -0.55° -0.41° -0.34° - 0.28° - 0.20° Total. Grams. 0.61 0.57 0.21 0.21 0.04 1.64 2. Intraperitoneal. The same dog received in this experiment 75.6 cc. of a dulse solu- tion containing 1.4 grams of pure substance. No reduction of Feh- ling's solution was observed before or after the injection. The time and rate of dulse excretion are shown in the following table: Examination of Urine. ROTATION. ESTIMATED EXCRETION OF DIILSE.* December 3, 2 P.M | December 3, 3 P.M i Injection December4, lOA.M | 133 December5, lOA.M I 200 December 5, 12 M ' 115 December 6 and 7 | 383 December 8, 10 A.M 520 December 9, 10 A.M 350 -0.14°t Grams. - 0.62° 0.36 - 0.52° 0.42 - 0.28° 0.05 -0.48° 0.69 -0.28 0.24 -0.20 Total 1.76 ficient hydrochloric acid to make the strength 2 per cent caused no subsequent production of sugar.- * All readings have been taken on the Ventzke scale, and calculated as angular degrees. t Estimating normal rotation of urine as — 0.17° (average). 336 Mary Davies Swartz, In both these experiments, the presence of dulse was readily detected by Fehling's solution in every urine which showed a high rotation. From the samples of the first 48 hours after injection, a considerable amount was isolated and identified as carbohydrate. It is evident that the excretion of this pentose-carbohydrate is gradual, commenc- ing soon after the injection, and continuing from four to five days. WTiile any quantitative estimate of the amount excreted, based on the changes in rotation, is subject to a high percentage of error, owing to normal fluctations in the rotation of the urine, as well as to analyt- ical discrepancies unavoidable in dealing with solutions containing only minute quantities of the substance under investigation, it is evi- dent that most of the dulse must have been excreted, and that, too, without any essential change in character. INJECTIONS or IRISH MOSS. 1. Subcutaneous. A dog weighing 9.4 kg. received 100 cc. of Irish moss solution, con- taining 1.5 grams of dry substance. No reducing substance occurred in the urine. Changes in rotation, due to the injection, are shown in the following table: Examination of Urine. VOLTTME. ROTATION. IRISH MOSS CC. -0.04° Injection 128 + 0.34° — 226 + 0.06° — 330 -0.20° 370 -0.14° May 18, 9 A.M. May 18, 4P.M. May 19, 9 A.M May 20, 9 A.M. May 21, 11A.M. May 22, 9 A.M Tests for Irish moss on May 19th were negative, but on May 20th- 22nd they were faintly positive. The experiment was discontinued at this point. The injection was not very well borne, the dog remain- ing lethargic throughout the period. 2. Intraperitoneal. Experiment A . A dog weighing 10 kg. received 160 cc. of an Irish moss solution containing 1.3 grams air dry material. Examination Nutrition Investigations. 337 for the presence of carbohydrate was made by testing the urine for reducing substances, before and after hydrolysis. The results are shown in the following table: Examination of Urine. REDUCTION OF FEHXING's SOLUTION. Before Hydrolysis. After Hydroly- OctoberlS, 11 A.M October 13, 12:30 P. M •. . . . Injection October 13, 2 P.M 27 October 13, 5 P.M 60 October 14, 9 A.M 450 October 15, 5 P.M 45 October 16, 9:30 A.M i — The urine before the injection showed a rotation of —0.14°, a sample of the mixed urines of October 13, 5 P.M., and October 14, 9 A.M., showed a rotation of —0.034°, the diminished levo-rota- tion undoubtedly due to the presence of this dextro-rotatory carbohy- drate. On hydrolysis, 50 cc. of this mixed sample yielded sugar equivalent to 0.035 grams of dextrose (by Allihn's method) . From the remainder of this sample, Irish moss carbohydrate was isolated; it formed a grayish- white powder, sweUing in water, and yielding mucic acid on oxidation with nitric acid. Experiment B. A dog weighing 9 kg. received intraperitoneally 100 cc. of a 2 per cent solution of Irish moss preparation. Examina- tion for carbohydrate was made as in the preceding experiments. The results appear in the following table: Examination of Urine. REDUCTION OF FEHUNG 'S SOLUTION. [Before Hydroly- After Hydroly- j sis. sis. October 30 1 P.M cc. Injection — 250 200 — 115 — October 30, 2:30 P.M _ October 31, 9 A.M November 1, 10 A.M -I- November 2, 10 A.M Irish moss was isolated and identified in the urine of November 1st. 338 Mary Davies Swartz, INJECTIONS OF SALEP. I. Subcutaneous. A dog weighing 7.2 kg. received 56 cc. of salep solution, containing 0.75 grams of pure mannan. No reducing substance was found in the urine. The changes in rotation, due to salep, are shown in the following table : Examination of Urine. TDSZ. VOLUME. ESTIMATION OF AMOUNT ROTATION. ! OF SALEP EXCRETED. May 17, CC. Injection 138 132 114 127 -0.17° - 0.27° -0.27° -0.20° -0.14° Grams. May 18 330 P.M May 19, 9 A.M May 20 9 A.M 0.3 May 21 9 A.M 0.3 May22, 9 A.M 0.04 May 22 5 P.M Salep was isolated and identified in the urines of May 20, 21, and 22. 2. Intraperitoneal. Experiment A . A dog weighing 7 kg. received 68 cc. of salep solu- tion, containing 1.2 grams of air dry mannan. No reducing substance was present in the urine at any time. Tests for the presence of salep by means of FeMing's solution, gave the following results: Examination of Urine. TIME. VOLUME. SALEP PRESENT. October 21, 12 M CC. Injection 125 190 October 21 2-30 P.M October 22, 9 A.M + October 23, 9 A.M 4- October 24, 9 A.M 140 — The salep was easily isolated and identified in the urine of October 22 and 23, the sugar obtained on hydrolysis being equivalent to 0.33 grams salep. Experiment B. A dog weighing 9.2 kg. received 80 cc. of salep so- Nutrition Investigations. 339 lution, containing 1.4 grams of air dry substance. No reducing sub- stance was detected in any of the urines. Tests for salep with Feh- ling's solution gave the following results: Examination of Urine. TIME. ' VOLUME. SAI^P PRESENT. October 24, 11 A.M October 24, 12 M October 25, 12 M 155 + October 26, 10 A.M 180 + October 27, 10 A.M 180 + October 28, 10 A.M From the urine of October 25, salep was isolated, which jdelded on hydrolysis 0.39 grams reducing sugar as dextrose; it was also isolated from the urines of the next two days, but was not estimated quanti- tatively. Experiment C. A dog weighing 9.2 kg. received 90 cc. of salep solution, containing 1.8 grams of pure mannan. No reduction of Fehling's solution occurred with any of the samples. Tests for salep with Fehling's solution gave the following results: Examination of Urine. ROTATION. SALEP PRESENT. December 2, 10 A.M i December 2, 2:30 P.M j Injection December 3, 10 A.M i 960 December 4, 10 A.M j 234 December 5, 10 A.M I 520 -0.17° -0.41° -0.27° + + (0.6gm.) + (0.5gm.) Unfortunately this experiment was imavoidably interrupted at this point. The salep was precipitated from 50 cc. of the urine for December 3, hydrolyzed, and sugar determined gra^dmetrically as dextrose, from which the total amount of salep in this day's urine was calculated as 0.67 gram. Salep determined in the same way on De- cember 4, showed an elimination of 0.18 gram; hence 0.85 gram was actually recovered in these two days. The influence of the levo- rotatory carbohydrate on the rotation of the urine was marked. 340 Mary Davies Swartz Experiment D. A dog weighing 6.4 kg. received 98 cc. of salep solution containing 1 gram of pure mannan. No reduction of Feh- ling's solution was observed throughout the experiment. The changes in rotation due to the salep are shown in the following table: Examination of Urine. ROTATION. salep precipitated by fehling's solution. January 31 February 1 February 2 February 3 Februar}' 4 February 5 116 Injection 152 238 154 137 -0.41° -0.41° -0.13° -0.13° - 0.20° + The results in this experiment are very puzzling. The normal rota- tion was high (—0.41°) for several weeks before this experiment but fairly constant, averaging —0.44°. If salep were excreted as mannan, the levo-rotation should have increased, yet it was decid- edly low on a day when salep was shown to be present, and also on a day when none could be detected. The absence of any positive tests for sugar, excluded the idea that the salep was being excreted in this form, but finally a sample of February 4, was tested with yeast, and marked fermentation observed. Unfortunately, this was after all the other samples had been discarded, hence no further tests could be made. Experiment E. A dog weighting 9.8 kg. received intraperitoneally 97.5 cc. of salep solution containing 1.3 grams pure mannan. No re- duction of Fehling 's solution was observed. The changes in rotation are shown in the first table on the next page. Salep was isolated and identified as carbohydrate, in the urines of May 19, 20, and 21, although the amount in the last two days was ap- parently too small to be detected by any change in the rotation. INJECTIONS OE SINISTRIN. I. Subcutaneous. A dog weighing 6.5 kg. received 49 cc. of sinistrin solution, contain- ing 3.3 grams pure substance. This solution showed a rotation of Nutrition Investigations. Examination of Urine. 341 ESTIMATION OF AMOUNT 01 SAIEP EXCRETED. Mayl7, 10A.M May 18, 10 A.M May 18, 3 P.M May 19, 9 A.M May 20, 9 A.M May 21, 11 A.M May 22, 9 A.M Injection 165 250 405 200 -0.14° -0.14° -0.34° - 0.14° -0.14° -0.14° Grams. 0.4 Salep present — pre- cipitated by Feh- ling's Solution. Salep present. No Salep present. -3.88° in a 200 mm. tube. The urine contained no reducing substance at any time. The changes in rotation, due to sinistrin in- jection, are shown in the following table: Examination of Urine. VOLUME. ESTIMATION or AMOUNT OB SINISTRIN EXCRETED. * January 15, 12 M.... January 15, 2:30 P.M January 16 January 17 and 18 . . . January 19 January 20 Injection 260 165 60 108 -0.41° -0.97° -0.41° -0.41° -0.47° Grams. 2.5 * CalcTilatlng f or sinistrin [a] d = —29.1°. 2. Intraperitoneal. Experiment A. A dog weighmg 6.5 kg. received 110 cc. of sinis- trin solution, containing 2 grams pure substance. This solution showed a rotation of - 1.18° in a 200 mm. tube. No reducing sub- stance was found in the urmes examined. The changes m rotation, due to sinistrin injection, are shown in the following table: 342 Mary Davies Swartz, Examination of Urine. ROTATION. ESTIMATION OF AMOUNT OF SINISTRXN EXCRETED.* January 11, 10:30 A.M January 11, 3 P.M January 12, 9:30 A.M January 13, 9:30 A.M January 14, 9:30 A.M * Calculating for slnlstrln \a\ d Injection 88 127 116 -0.48° - 2.04° - 0.48° -0.48° Grams. 2.7 = —29.1°. Experiment B. A dog weighing 4.6 kg. received 108 cc. of sinis- trin solution, containing 2.3 grams pure substance. The rotation of this solution was —1.38" in a 200 mm. tube. No reducing sub- stance was detected in the urine at any time. The changes in rota- tion are shown in the following table: Examination of Urine. TIME. VOLUME. ROTATION. ESTIMATION OF AMOUNT OF SINISTRIN EXCRETED.* January 26 CC. Injection 148 95 155 -0.14° - 1.38° -0.41° -0.14° Grams. January 27, 9:30 A.M January 27, 5 :P.M January 28, 9 :AM 2.1 0.4 January 29, 9 :A.M ■ Calculating for slnlstrln [a] d = — 29.1° In all these experiments, the sinistrin was isolated and identified as a levo-rotatory carbohydrate, yielding reducing sugar on hydrolysis. It was apparently excreted quantitatively in every case. FEEDING EXPERIMENTS. Methods and Technique. Feeding experiments were conducted with dogs and human sub- jects, under conditions as nearly normal as possible. The dogs were kept in metal cages, arranged for the separate collection of urine and faeces. They were fed once a day, on a uniform weight diet, consist- ing of chopped lean meat, lard, and cracker meal, in suitable portions Nutrition Investigations. 343 and amounts to maintain a constant body weight. The carbohydrate under investigation was dissolved or suspended in water, and mixed with this basal ration. In the earlier experiments the periods were divided as follows: Fore = 3 days on the basal ration; mid = 3 days in which some preparation was added, the amoimt being the same each day; after = 3 days like the fore period. Separation of the pe- riods in the faeces was accomplished by marking with soot or carmine capsules. In all later experiments, two days constituted the fore pe- riod, and a day on the normal diet was included at the beginning and end of the mid period, making thus four days, to insure against any of the material under investigation being carried into the faeces of the after period. In several cases, the presence or absence of galactans or mannans in the faeces has been verified by testing the hydrolyzed material for mucic acid or mannose-hydrazone. For analysis, the faeces, collected and weighed, were rubbed to a thin mud with alcohol, dried to constant weight on a water bath, weighed air dry, and ground finely in a coffee mill. The portions constituting each period were thoroughly mixed, and from 2 to 5 grams taken for hydrolysis, according to the yield of carbohydrate anticipated. The samples were boiled on a reflex condenser with 100 cc. of 2 per cent hydrochloric acid, for two hours; or longer if thought to contain a carbohydrate which previous analysis^ had shown to require more time for complete hydrolysis. The products of hydrolysis, cooled and neutralized, were made up to 250 cc. and sugar determined as dextrose by Allihn's gravimetric method. It was found that the copper reduction was often very in- complete, and that much more satisfactory results came from clari- fying the solutions with charcoal after making up to volume. Not only were duplicate analyses in closer agreement, but in some cases the yield of cupric oxide was two or three times greater than before this treatment. Owing to the complexity and diversity of the prod- ucts of hydrolysis, results are at best only approximate. In experiments with dulse, the pentosans were determined by the phloroglucin method. ^ The human subjects were healthy, active young women. Their diet was not weighed, but was kept as uniform as possible. All cel- 1 Cf. table, p. 317. *Cf. Official and Provisional Methods of Analysis, Bulletin No. 107 (1907), Bureau of Chemistry, United States Department of Agriculture. 344 Mary Davies Swartz, lulose-containing foods, such as nuts, frmts, green vegetables, peas and beans, coarse bread and cereals, were carefully avoided; so that the carbohydrates were limited almost entirely to bread and crackers made from fine white flour, a small quantity of potato, and sugar. To this diet the gelatinizing carbohydrates were added in the form of blanc mange or jelly; dulse was dissolved in some beverage, and the insoluble preparations boiled half an hour in a little water and eaten as a vegetable, seasoned with salt, butter, and vinegar. The blanc manges or jellies made from the Hawaiian seaweed preparations were equally attractive in texture and flavor with those made from Irish moss. Periods were marked, and the analyses of faeces conducted in the manner already described for the experiments with dogs. The Digestibility of Pentosans. Four preparations were fed. Dulse, ^ Limu Eleele,^ Limu Lipoa,^ and Limu Pahapaha,^ without production of unpleasant symptoms in any case. The results of all trials are shown in the tables on the following pages. 1 Cf. p. 303. 2 Cf. p. 307. 3 Cf. p- 308. g@ 8 P4 o: s S tP CO a c^ CO 1-1 > « D o •^ O A ^ kO '^ i d -* (N (N CO 7—1 I> t-i m Q O. — E4 i «o 1— 1 lO o ■3 tn be 1 t5 CO d >* 00 CD 1— 1 d CO d d 00 d lO (M C -*! th 2 < K (T^ oi 00 ,—1 d ,— i ,_J c CD CO d 1—1 I— 1 O S 2 2 CD 00 Ti< o 05 lO i ^ U (M (M 1— 1 1—1 a a. Q, a, (U 0) . y T3 ^ "^ tn >> tn "^ tn ^ rt c3 rt sj d cS S^ c3 CS oj C!3 73 T3 Tj -d T3 X) -d •T3 x) t3 -0 d CO CO CO C<( ■* (M (M Tf< (M K 3 8 (^ a '~- w g s CO -<# •* 1 CO lo "3 si ►J p m u s 1> -* o C^ 1—1 CO 00 CO « "s g o CO o 1—1 t-^ 1-1 ^ '^ 1—1 fa O 11 o t* (i ^ H ^ -* o CO 00 00 02 -* ^ S (J^ O o H S w >< a lO o lO CO c^ CO t^ (M 10 3 < o 1 IC CO Tt^ ^ 00 t^ 1—1 C5 (M O T— 1 CO 1—1 7—1 00 1— ( Tt< 00 (N H . W H o E2 a o CO CO CO d t^ CO d CO t^ 3 o CO CO c^ CO (M (M CO g a "i (M 1-H CO X) 'O -d a; > I— 1 S 1 bO hO O O '■'— ^ Q Q N 6 lO CO l> saraas < < <1 Q H ^ g « s o § > o ^^ OS O 1 s« i (^ o g §<- Q) ^ o ^ so CO rH HJ >«. T— 1 CO . >. >, rt c^ 5^ T3 -d -a (N •* IM II II II ^ ^ -d i ^ 1 (M 7—1 1— 1 > be O C ■; " < 1 348 Mary Davies Swartz, The coefScients of digestibility of the pentosan preparations, as determined in the usual way from the preceding experiments, are set forth in the following table: SERIES A. PENTOSAN. COEFFICIENT OF DIGESTIBIUTY. EXPERIMENT NO. For the Dog. For Man. 1 Dulse 80 2 Dulse 66 3 Dulse 100 4 Dulse 100 5 Limu Eleele 50 6 Limu Eleele 20 7 Limu Eleele 69 8 Limu Pahapaha 34 9 Limu Lipoa 16 It is evident from these figures, that pentosans in soluble form dis- appear from the alimentary tract of dogs to a very considerable extent (average 73 per cent), and that small quantities, ingested by man, do not reappear in the faeces. The insoluble limu preparations appear much more indigestible, an average of 28 per cent being digested by dogs, and 51 per cent by man. It must be borne in mind, in interpreting the results of these metabo- lism experiments, that they are at best only approximate. The dif- ficulty of strict separation of the faeces, the fact that the human sub- jects were not kept on a uniform weighed diet, and the errors unavoid- ably introduced by determining many different kinds of sugar as dex- trose, make all of the figures given as "coefiicients of digestibility," in this and succeeding sections, comparative rather than absolute. The Digestibility of Galactans. In these experiments, preparations of the water extracts of Irish moss, Limu Manauea, Limu Huna and Limu Akiaki have been fed, without any disagreeable symptoms. The results are given in the tables which follow: Q W ^ g ■* o Oi o ^ lO 00 00 o 3 o K 1—1 o (^ Pi IMU FED As Dex- trose). i 00 05 00 1— c 1-1 M i-i ^-^ < en i t^ 00 Oi Tj< 00 CO 00 CD IM o O CO t^ "'i »^ (N d IM ,-!) t- T-i d ^ CO (M o ^ § 11 §1 ^ " .—1 CO 2 o H PQ 01 8 -* "3 o CD Tf< o ■* 00 t-H t^ O CD O g o O ».. Ci >o as 00 CD CD d ^ d »o 1—1 o> ftj CO (M >— 1 c^ * * .^ ++ u 1 T-H 23 id C5 CD t^ t^ O lO 00 00 00 >o CO O CD lO Tjl ^ a )>.. (M Tt* I— 1 i-H CD Cl -H C5 CO CO C5 CO ^ § (i g « i 05 05 t^ 00 d CD O t^ CO 3 9 g lO lO lO CD CO ^ a o Tfl ^ ^"^ CO 1-4 OT in in tn tn CO tn tn O O O O a . a a a i « S 'm 43 .SS as x) _2 bo"C X3 CO I-I _o •C X 1— 1 _o O tn 't^ O . 'C ';-* '^ 2 S 0 C/3 -^^ CO c -' 4- 7^ aj o O xT ^ CD tn "-' C!3 o si . >. >> >> >. >^ >^ !>. >> >% >> nj a b3 03 rt rt rt rt rt cS 03 o3 XI X! Td X) XJ -d -^ T3 X) X) -d X) e CO (M CO CO CO CO (M ^ (M CO -* CO 2 II II 11 II II 11 II II II 11 II II ^ XJ 5 X) aj o Li X) li fi X) 5 o ■^ 1 o i 1^ i < o ■^ < ^' ^ ^ O -* o o .^ 00 d «m' CO M 1— ( -* »o & ^.^ i-i P I-I 1— 1 bO O P 03 a o o3 a 1 7 6 1—1 (N CO -* • £; , ^i pq M pq PQ P c K -i^ 2 w § O 00 o o o S > **^ ■* CO t- o -* 3 o K (^ B! s CO CO '^ CD S^i ;S H ►J ^^ W UJ s (M CD CD t^ o O tH CO Tt< 00 00 GO CO (S 2 ^ lO ^ d "* !-< (M Oi 1— H T-H l>^ -^ 1—1 o 11 (^ 2 >- XI O gp H O g o V. CO CO O GO O C2 GO iO ^ 1-1 U3 CD 1—1 1— ( 00 CO CO Tfi 00 '-I d <^ .— 1 T— 1 rH O H ><■ CO o CO lo ^- tH O Cft (N It- tV 02 C5 B d O Q 1 '^i f ( i>i CO -^ oi d CO tv 1—1 d t- ■^ 1 e S CO l-H CO ,-H Ttl 00 CO ^ Ci CD (M Eh . CO CO CO o t^ 00 as t^ ^ 02 CO lo 00 K E- o 2 a o M< Ir- Tt^ »o lo 00 t^ d d CD (N o M 2 CO Ci iM CO ca CO >o as lO »o CO rfH 05 ^ 1— ( (M T— 1 1—1 liti (N cd ci c^ •k di S S § S S ;3 1 S 1^ 3 13 «5 g g a a .2 t3 o 6 :^ « ;3 § ;3 g ;3 § .2 s 3 O tn +3 cS u i ^ meal, 30 10 gms. preparati ^ 1 Ph gms. parati gms. parati 4) be g O M o S 2 - 3 ^ o fa «5 O <>5 § ^ + CIj ^..^ cS o tn 4_r M % >> >. >. >, >. >. >^ >. >> >> >> >i cS rt cvi rt C^ rt rt irf rf rt rf i2 ce -3 'O -^ TJ TS -n -o "d -Td -O ■S -« -^ d (M ■* C<) C^ Tt< (M (N lO C^ (M lO "* (M 2 II II 11 II II II II II II II II II II ;-4 'O 1-1 'Td -xi Td -o O i <^ fS i ^tL i i i '' "^^ ^ H o (M w o « CO 1—1 CO (N w fc«! 7—1 l-H lO lO ^ ^—^ ^^ c c rt cj E. > 1— 1 bJD O P > P a o a o 6 lO CD t- GO z 1 oi P3 m pq pq d H •h» S o t- O K CO 05 ftj ca § d^ ^ ^' t^ t^ Oi o (^ —' O b H CO Oi CO ^ lO lO (M 00 1> rt pj i O ft •*' CO i>^ 03 d lO 00 '^ CO U g v.. »o ^ t— ( CO 05 ^ >. >. >. b >. b >> >> cS c^ cS ci c^S rt cd cd Gj -d Id T3 -n TJ Xf XJ -o ^3 i (N ^ (M (M ^ (M C^ lO (M II II 11 11 II 11 II II II -n X) 'tS O i ^ o % _< ^ _f^ ^ ^ H o (M n a ^ CO ,_) CO H 1—1 T— I -* ^ ^_^ S3 cS > 1— 1 > a 1 P bD bo O P O X d C2 o ^ 2 ^ pq M w O ^ (M t- CO O lO ""^ CO (M o CO ^ CO 00 O I-H »o + 3 fe O crt 3 ri^ b < a rrl CS u CAJ O! tn rn >. >. >> cj at oj TJ -d Td CO < 9, fe <^ Bj o , g g^ ^ O o t- i> Sp s '^ "*' (N (M S « 2 S CO CO (M C<) ii P4 o r< w ^ a t^ Ol (M r- »o IM O lO I-H 00 j>- CD tK g s 2 d ^ ^ (N lO CO 0 00 "# d d d (^ CO T-H T— 1 u ^- Tf< lO O 1— 1 (M 00 CO i-i Tt< lO IV I-H 1 CO ^ CD Oi d r-l »0 CO 02 05 ^ t^ ^ CO o lO CD CD ^ 3 > >. >> >. >. >> >^ >> >. >. >s a oi rt k5 d Ki rf n3 d o3 rt cd Xi -d -a •n -0 X) -d -d -d T3 TJ XJ Q O Pi CO CO CO CO CO CO (M ^ c^ (N ^ (N II II II II II II II II II II II II 1 txl v-i M J-( >-< fM -d t^ 03 C<) (M iS" lO ■* & '5' 'c" 1 B 1— 1 o 6 M f3 be 1— 1 bO 1 t o O 1 Q P N X i 6 ^ (M CO "* ~\ ^; 1 ui CJ u U o " 2 i H H o a Ot « a H a "to o ►J > o [ 3 8 fe (j; 1 « i o , ' o O o SALEP (AsD trosi ci GO 00 ci 1—1 I f^ c ! i=3 < w 3 o CO t^ >o (M CO lO o ^ i f^ w ^- 5 H ^ >^ T-H o o T-H -* (M C^ CO y—i 1 O S £ ^ £5 O 8 « go 1 O o a GO «3 O o CO CO CO o CO CO u * g GO CO CO 00 o 1> '-J CO IM O O a 1—1 ^ CO Tt< (^ t^ lO tj (M 05 oq c C! h cd cd nj ^ ^ ^ «5 R- a a, -o _!> -c (D t3 _o "rt _o "rf _o *n CO *n in Ut ■^^ ^ > >. >. >^ >> >> >> >. >, rt rt rf c3 p3